# SPDX-License-Identifier: Apache-2.0 # SPDX-FileCopyrightText: Copyright contributors to the vLLM project # Copyright 2025 The Qwen team. # Copyright 2023 The vLLM team. # Copyright 2022 EleutherAI and the HuggingFace Inc. team. All rights reserved. # # This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX # and OPT implementations in this library. It has been modified from its # original forms to accommodate minor architectural differences compared # to GPT-NeoX and OPT used by the Meta AI team that trained the model. # # 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 Qwen3-Omni-Moe model (thinker part).""" from collections.abc import Callable, Iterable, Iterator, Mapping, Sequence from functools import partial from typing import Any, Literal, cast import numpy as np import torch import torch.nn as nn import torch.nn.functional as F from packaging.version import Version from transformers.feature_extraction_utils import BatchFeature from transformers.models.qwen3_omni_moe.configuration_qwen3_omni_moe import ( Qwen3OmniMoeAudioEncoderConfig, Qwen3OmniMoeConfig, Qwen3OmniMoeThinkerConfig, ) from transformers.models.qwen3_omni_moe.processing_qwen3_omni_moe import ( Qwen3OmniMoeProcessor, ) from transformers.models.whisper import WhisperFeatureExtractor # isort: off from transformers import PretrainedConfig from transformers import __version__ as TRANSFORMERS_VERSION # isort: on from vllm.compilation.decorators import support_torch_compile from vllm.config import ModelConfig, SpeechToTextConfig, VllmConfig from vllm.distributed import get_pp_group, get_tensor_model_parallel_world_size from vllm.inputs.data import PromptType from vllm.logger import init_logger from vllm.model_executor.layers.activation import _ACTIVATION_REGISTRY from vllm.model_executor.layers.attention.mm_encoder_attention import ( MMEncoderAttention, ) from vllm.model_executor.layers.conv import Conv3dLayer from vllm.model_executor.layers.linear import ( ColumnParallelLinear, QKVParallelLinear, RowParallelLinear, ) 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 from vllm.model_executor.model_loader.weight_utils import default_weight_loader from vllm.model_executor.models.module_mapping import MultiModelKeys from vllm.model_executor.models.qwen2_audio import Qwen2AudioProcessingInfo from vllm.multimodal import MULTIMODAL_REGISTRY from vllm.multimodal.inputs import MultiModalFeatureSpec, MultiModalKwargsItems from vllm.multimodal.parse import AudioProcessorItems, MultiModalDataItems from vllm.multimodal.processing.processor import ( MultiModalPromptUpdates, PlaceholderFeaturesInfo, PromptReplacement, PromptUpdate, PromptUpdateDetails, ) from vllm.sequence import IntermediateTensors from vllm.transformers_utils.processor import cached_processor_from_config from vllm.v1.attention.backends.registry import AttentionBackendEnum from .interfaces import ( MultiModalEmbeddings, SupportsMRoPE, SupportsMultiModal, SupportsPP, SupportsTranscription, ) from .qwen2_5_omni_thinker import ( Qwen2_5OmniAudioFeatureInputs, Qwen2_5OmniConditionalGenerationMixin, Qwen2_5OmniThinkerDummyInputsBuilder, Qwen2_5OmniThinkerMultiModalProcessor, check_interleaved_audio_video, merge_interleaved_embeddings, ) from .qwen2_5_vl import ( Qwen2_5_VisionAttention, Qwen2_5_VLProcessingInfo, ) from .qwen3_moe import Qwen3MoeForCausalLM, Qwen3MoeModel from .utils import ( AutoWeightsLoader, WeightsMapper, _merge_multimodal_embeddings, maybe_prefix, ) from .vision import get_vit_attn_backend logger = init_logger(__name__) # Speech input languages supported by Qwen3-Omni # From: https://huggingface.co/Qwen/Qwen3-Omni-30B-A3B-Instruct ISO639_1_SUPPORTED_LANGS = { "en": "English", "zh": "Chinese", "ko": "Korean", "ja": "Japanese", "de": "German", "ru": "Russian", "it": "Italian", "fr": "French", "es": "Spanish", "pt": "Portuguese", "ms": "Malay", "nl": "Dutch", "id": "Indonesian", "tr": "Turkish", "vi": "Vietnamese", "yue": "Cantonese", "ar": "Arabic", "ur": "Urdu", } def _get_feat_extract_output_lengths(input_lengths: torch.Tensor): input_lengths_leave = input_lengths % 100 feat_lengths = (input_lengths_leave - 1) // 2 + 1 output_lengths = ( ((feat_lengths - 1) // 2 + 1 - 1) // 2 + 1 + (input_lengths // 100) * 13 ) return output_lengths # ============= Audio Encoder Components ============= class SinusoidsPositionEmbedding(nn.Module): """Sinusoidal position embedding for audio encoder.""" def __init__(self, length: int, channels: int, max_timescale: int = 10000): super().__init__() self.length = length self.channels = channels self.max_timescale = max_timescale if channels % 2 != 0: raise ValueError("SinusoidsPositionEmbedding needs even channels input") log_timescale_increment = np.log(max_timescale) / (channels // 2 - 1) inv_timescales = torch.exp( -log_timescale_increment * torch.arange(channels // 2).float() ) scaled_time = ( torch.arange(length)[:, np.newaxis] * inv_timescales[np.newaxis, :] ) positional_embedding = torch.cat( [torch.sin(scaled_time), torch.cos(scaled_time)], dim=1 ) self.register_buffer( "positional_embedding", positional_embedding, persistent=False ) def forward(self, seqlen: int) -> torch.Tensor: return self.positional_embedding[:seqlen, :] class Qwen3OmniMoeAudioAttention(nn.Module): """Multi-headed attention for Qwen3-Omni Audio Encoder using MMEncoderAttention.""" def __init__( self, config: Qwen3OmniMoeAudioEncoderConfig, prefix: str = "", ): super().__init__() self.embed_dim = config.d_model self.num_heads = config.encoder_attention_heads self.head_dim = self.embed_dim // self.num_heads tp_size = get_tensor_model_parallel_world_size() self.num_local_heads = self.num_heads // tp_size if (self.head_dim * self.num_heads) != self.embed_dim: raise ValueError( f"embed_dim must be divisible by num_heads (got `embed_dim`: " f"{self.embed_dim} and `num_heads`: {self.num_heads})." ) self.scaling = self.head_dim**-0.5 self.qkv = QKVParallelLinear( hidden_size=self.embed_dim, head_size=self.head_dim, total_num_heads=self.num_heads, total_num_kv_heads=self.num_heads, bias=True, prefix=f"{prefix}.qkv", ) self.out_proj = RowParallelLinear( input_size=self.embed_dim, output_size=self.embed_dim, bias=True, prefix=f"{prefix}.out_proj", ) self.attn = MMEncoderAttention( num_heads=self.num_local_heads, head_size=self.head_dim, scale=self.scaling, prefix=f"{prefix}.attn", ) def forward( self, hidden_states: torch.Tensor, cu_seqlens: torch.Tensor, max_seqlen: torch.Tensor | None, ) -> torch.Tensor: seq_length, _ = hidden_states.size() qkv, _ = self.qkv(hidden_states) q, k, v = qkv.chunk(3, dim=-1) q = q.view(1, seq_length, -1, self.head_dim) k = k.view(1, seq_length, -1, self.head_dim) v = v.view(1, seq_length, -1, self.head_dim) attn_output = self.attn( query=q, key=k, value=v, cu_seqlens=cu_seqlens, max_seqlen=max_seqlen, ) attn_output = attn_output.view(seq_length, -1) output, _ = self.out_proj(attn_output) return output class Qwen3OmniMoeAudioEncoderLayer(nn.Module): """Transformer encoder layer for Qwen3-Omni Audio Encoder.""" def __init__( self, config: Qwen3OmniMoeAudioEncoderConfig, prefix: str = "", ): super().__init__() self.embed_dim = config.d_model self.self_attn = Qwen3OmniMoeAudioAttention( config, prefix=f"{prefix}.self_attn" ) self.self_attn_layer_norm = nn.LayerNorm(self.embed_dim) self.activation_fn = _ACTIVATION_REGISTRY[config.activation_function] self.fc1 = ColumnParallelLinear( self.embed_dim, config.encoder_ffn_dim, bias=True, prefix=f"{prefix}.fc1", ) self.fc2 = RowParallelLinear( config.encoder_ffn_dim, self.embed_dim, bias=True, prefix=f"{prefix}.fc2", ) self.final_layer_norm = nn.LayerNorm(self.embed_dim) def forward( self, hidden_states: torch.Tensor, cu_seqlens: torch.Tensor, max_seqlen: torch.Tensor | None, ) -> torch.Tensor: """ Args: hidden_states: Input tensor of shape (seq_len, hidden_size) cu_seqlens: Cumulative sequence lengths max_seqlen: Maximum sequence length in the batch """ residual = hidden_states hidden_states = self.self_attn_layer_norm(hidden_states) hidden_states = self.self_attn( hidden_states=hidden_states, cu_seqlens=cu_seqlens, max_seqlen=max_seqlen, ) hidden_states = residual + hidden_states residual = hidden_states hidden_states = self.final_layer_norm(hidden_states) hidden_states, _ = self.fc1(hidden_states) hidden_states = self.activation_fn(hidden_states) hidden_states, _ = self.fc2(hidden_states) hidden_states = residual + hidden_states # Clamp for numerical stability with fp16 if hidden_states.dtype == torch.float16: clamp_value = torch.finfo(hidden_states.dtype).max - 1000 hidden_states = torch.clamp( hidden_states, min=-clamp_value, max=clamp_value ) return hidden_states class Qwen3OmniMoeAudioEncoder(nn.Module): """vLLM-native Qwen3-Omni Audio Encoder.""" def __init__( self, config: Qwen3OmniMoeAudioEncoderConfig, prefix: str = "", ): super().__init__() embed_dim = config.d_model self.num_mel_bins = config.num_mel_bins self.max_source_positions = config.max_source_positions self.n_window = config.n_window self.n_window_infer = config.n_window_infer self.conv_chunksize = config.conv_chunksize # Position embedding self.positional_embedding = SinusoidsPositionEmbedding( self.max_source_positions, embed_dim ) # Convolutional layers for mel-spectrogram processing self.conv2d1 = nn.Conv2d(1, config.downsample_hidden_size, 3, 2, padding=1) self.conv2d2 = nn.Conv2d( config.downsample_hidden_size, config.downsample_hidden_size, 3, 2, padding=1, ) self.conv2d3 = nn.Conv2d( config.downsample_hidden_size, config.downsample_hidden_size, 3, 2, padding=1, ) conv_out_dim = config.downsample_hidden_size * ( (((config.num_mel_bins + 1) // 2 + 1) // 2 + 1) // 2 ) self.conv_out = nn.Linear(conv_out_dim, config.d_model, bias=False) # Transformer encoder layers self.layers = nn.ModuleList( [ Qwen3OmniMoeAudioEncoderLayer( config, prefix=f"{prefix}.layers.{i}", ) for i in range(config.encoder_layers) ] ) # Output layers self.ln_post = nn.LayerNorm(config.d_model) self.proj1 = nn.Linear(config.d_model, config.d_model) self.act = _ACTIVATION_REGISTRY[config.activation_function] self.proj2 = nn.Linear(config.d_model, config.output_dim) # Get attention backend self.attn_backend = get_vit_attn_backend( head_size=config.d_model // config.encoder_attention_heads, dtype=torch.get_default_dtype(), ) def compute_attn_mask_seqlen(self, cu_seqlens: torch.Tensor) -> torch.Tensor | None: """Compute max_seqlen only for flash attention backends.""" max_seqlen = None if self.attn_backend in { AttentionBackendEnum.FLASH_ATTN, AttentionBackendEnum.ROCM_AITER_FA, AttentionBackendEnum.TRITON_ATTN, }: max_seqlen = (cu_seqlens[1:] - cu_seqlens[:-1]).max() return max_seqlen @property def dtype(self) -> torch.dtype: return self.conv2d1.weight.dtype @property def device(self) -> torch.device: return self.conv2d1.weight.device def forward( self, input_features: torch.Tensor, feature_lens: torch.Tensor, aftercnn_lens: torch.Tensor, ): # Compute chunk information chunk_num = torch.ceil(feature_lens / (self.n_window * 2)).long() chunk_lengths = torch.tensor( [self.n_window * 2] * chunk_num.sum(), dtype=torch.long, device=feature_lens.device, ) tail_chunk_index = F.pad(chunk_num, (1, 0), value=-1).cumsum(0)[1:] chunk_lengths[tail_chunk_index] = feature_lens % (self.n_window * 2) chunk_lengths[chunk_lengths == 0] = self.n_window * 2 # Split input features into chunks and pad chunk_list = input_features.T.split(chunk_lengths.tolist(), dim=0) padded_feature = nn.utils.rnn.pad_sequence( chunk_list, batch_first=True ).transpose(1, 2) # Compute feature lengths after CNN feature_lens_after_cnn = self._get_cnn_output_lengths(chunk_lengths) # Vectorized mask creation: avoid creating many small tensors max_len_after_cnn = feature_lens_after_cnn.max().item() indices = torch.arange(max_len_after_cnn, device=padded_feature.device) padded_mask_after_cnn = indices.unsqueeze(0) < feature_lens_after_cnn.unsqueeze( 1 ) # Add channel dimension for conv2d padded_feature = padded_feature.unsqueeze(1) # Apply convolutional layers (chunk if needed to avoid OOM) if padded_feature.size(0) <= self.conv_chunksize: # Fast path: no chunking needed padded_embed = F.gelu(self.conv2d1(padded_feature)) padded_embed = F.gelu(self.conv2d2(padded_embed)) padded_embed = F.gelu(self.conv2d3(padded_embed)) else: # Chunked processing to avoid OOM padded_embeds = [] for chunk in padded_feature.split(self.conv_chunksize, dim=0): padded_embed = F.gelu(self.conv2d1(chunk)) padded_embed = F.gelu(self.conv2d2(padded_embed)) padded_embed = F.gelu(self.conv2d3(padded_embed)) padded_embeds.append(padded_embed) padded_embed = torch.cat(padded_embeds, dim=0) # (batch, channels, freq, time) -> (batch, time, channels*freq) b, c, f, t = padded_embed.size() padded_embed = self.conv_out( padded_embed.permute(0, 3, 1, 2).contiguous().view(b, t, c * f) ) # Add positional embedding positional_embedding = ( self.positional_embedding.positional_embedding[: padded_embed.shape[1], :] .unsqueeze(0) .to(padded_embed.dtype) ) padded_embed = padded_embed + positional_embedding # Extract valid hidden states and compute cu_seqlens hidden_states = padded_embed[padded_mask_after_cnn] # Compute cumulative sequence lengths for chunked attention cu_chunk_lens = [0] window_aftercnn = padded_mask_after_cnn.shape[-1] * ( self.n_window_infer // (self.n_window * 2) ) # Use tolist() for efficient batch conversion from tensor to Python for cnn_len in aftercnn_lens.tolist(): num_full_chunks = cnn_len // window_aftercnn remainder = cnn_len % window_aftercnn cu_chunk_lens.extend([window_aftercnn] * num_full_chunks) if remainder: cu_chunk_lens.append(remainder) cu_seqlens = torch.tensor(cu_chunk_lens, device=aftercnn_lens.device).cumsum( -1, dtype=torch.int32 ) max_seqlen = self.compute_attn_mask_seqlen(cu_seqlens) # Apply transformer layers for encoder_layer in self.layers: hidden_states = encoder_layer( hidden_states, cu_seqlens, max_seqlen, ) # Apply output layers hidden_states = self.ln_post(hidden_states) hidden_states = self.proj1(hidden_states) hidden_states = self.act(hidden_states) hidden_states = self.proj2(hidden_states) return hidden_states def _get_cnn_output_lengths(self, input_lengths: torch.Tensor) -> torch.Tensor: """Compute output lengths after the three conv2d layers.""" lengths = input_lengths for _ in range(3): lengths = (lengths - 1) // 2 + 1 return lengths def load_weights(self, weights: Iterable[tuple[str, torch.Tensor]]) -> set[str]: """Load weights with mapping from HuggingFace format.""" stacked_params_mapping = [ # (param_name, shard_name, shard_id) ("self_attn.qkv.", "self_attn.q_proj.", "q"), ("self_attn.qkv.", "self_attn.k_proj.", "k"), ("self_attn.qkv.", "self_attn.v_proj.", "v"), ] params_dict = dict(self.named_parameters(remove_duplicate=False)) loaded_params: set[str] = set() for name, loaded_weight in weights: for param_name, weight_name, shard_id in stacked_params_mapping: if weight_name not in name: continue name = name.replace(weight_name, param_name) param = params_dict[name] weight_loader = param.weight_loader weight_loader(param, loaded_weight, shard_id) break else: param = params_dict.get(name) if param is not None: weight_loader = getattr( param, "weight_loader", default_weight_loader ) weight_loader(param, loaded_weight) loaded_params.add(name) return loaded_params class Qwen3_VisionPatchEmbed(nn.Module): def __init__( self, patch_size: int = 14, temporal_patch_size: int = 2, in_channels: int = 3, hidden_size: int = 1152, ) -> None: super().__init__() self.patch_size = patch_size self.temporal_patch_size = temporal_patch_size self.hidden_size = hidden_size kernel_size = (temporal_patch_size, patch_size, patch_size) self.proj = Conv3dLayer( in_channels, hidden_size, kernel_size=kernel_size, stride=kernel_size, bias=True, ) def forward(self, x: torch.Tensor) -> torch.Tensor: L, _ = x.shape x = x.view(L, -1, self.temporal_patch_size, self.patch_size, self.patch_size) x = self.proj(x).view(L, self.hidden_size) return x class Qwen3_VisionMLP(nn.Module): def __init__( self, in_features: int, hidden_features: int, bias: bool = False, act_fn: Callable[[torch.Tensor], torch.Tensor] = F.silu, quant_config: QuantizationConfig | None = None, prefix: str = "", ): super().__init__() self.linear_fc1 = ColumnParallelLinear( in_features, hidden_features, bias=bias, quant_config=quant_config, return_bias=False, prefix=f"{prefix}.linear_fc1", ) self.linear_fc2 = RowParallelLinear( hidden_features, in_features, bias=bias, quant_config=quant_config, return_bias=False, prefix=f"{prefix}.linear_fc2", ) self.act_fn = act_fn def forward(self, x: torch.Tensor): mlp_output = self.linear_fc2(self.act_fn(self.linear_fc1(x))) return mlp_output class Qwen3_VisionBlock(nn.Module): def __init__( self, dim: int, num_heads: int, mlp_hidden_dim: int, act_fn: Callable[[torch.Tensor], torch.Tensor] = F.silu, norm_layer: Callable[[int], nn.Module] | None = None, quant_config: QuantizationConfig | None = None, prefix: str = "", ) -> None: super().__init__() if norm_layer is None: norm_layer = partial(nn.LayerNorm, eps=1e-6) self.norm1 = norm_layer(dim) self.norm2 = norm_layer(dim) self.attn = Qwen2_5_VisionAttention( embed_dim=dim, num_heads=num_heads, projection_size=dim, quant_config=quant_config, prefix=f"{prefix}.attn", ) self.mlp = Qwen3_VisionMLP( dim, mlp_hidden_dim, act_fn=act_fn, bias=True, quant_config=quant_config, prefix=f"{prefix}.mlp", ) def forward( self, x: torch.Tensor, cu_seqlens: torch.Tensor, rotary_pos_emb_cos: torch.Tensor, rotary_pos_emb_sin: torch.Tensor, max_seqlen: torch.Tensor | None, # Only used for Flash Attention ) -> torch.Tensor: x = x + self.attn( self.norm1(x), cu_seqlens=cu_seqlens, rotary_pos_emb_cos=rotary_pos_emb_cos, rotary_pos_emb_sin=rotary_pos_emb_sin, max_seqlen=max_seqlen, ) x = x + self.mlp(self.norm2(x)) return x class Qwen3_VisionPatchMerger(nn.Module): def __init__( self, d_model: int, context_dim: int, norm_layer: Callable[[int], nn.Module] | None = None, spatial_merge_size: int = 2, use_postshuffle_norm: bool = False, quant_config: QuantizationConfig | None = None, prefix: str = "", ) -> None: super().__init__() self.hidden_size = context_dim * (spatial_merge_size**2) self.use_postshuffle_norm = use_postshuffle_norm if self.use_postshuffle_norm: context_dim = self.hidden_size if norm_layer is None: norm_layer = partial(nn.LayerNorm, eps=1e-6) self.use_postshuffle_norm = use_postshuffle_norm self.ln_q = norm_layer( self.hidden_size if use_postshuffle_norm else context_dim ) self.mlp = nn.ModuleList( [ ColumnParallelLinear( self.hidden_size, self.hidden_size, bias=True, quant_config=quant_config, prefix=f"{prefix}.mlp.0", ), nn.GELU(), RowParallelLinear( self.hidden_size, d_model, bias=True, quant_config=quant_config, prefix=f"{prefix}.mlp.2", ), ] ) def forward(self, x: torch.Tensor) -> torch.Tensor: if self.use_postshuffle_norm: x = self.ln_q(x.view(-1, self.hidden_size)) else: x = self.ln_q(x).view(-1, self.hidden_size) mlp_fc1, mlp_act, mlp_fc2 = self.mlp x_parallel, _ = mlp_fc1(x) x_parallel = mlp_act(x_parallel) out, _ = mlp_fc2(x_parallel) return out class Qwen3Omni_VisionTransformer(nn.Module): def __init__( self, vision_config, norm_eps: float = 1e-6, quant_config: QuantizationConfig | None = None, prefix: str = "", ) -> None: super().__init__() self.hidden_size = vision_config.hidden_size self.num_heads = vision_config.num_heads self.image_size = vision_config.image_size self.patch_size = vision_config.patch_size self.spatial_merge_size = vision_config.spatial_merge_size self.spatial_merge_unit = self.spatial_merge_size**2 self.temporal_patch_size = vision_config.temporal_patch_size self.num_grid_per_side = self.image_size // self.patch_size self.apply_vit_abs_pos_embed = vision_config.apply_vit_abs_pos_embed self.deepstack_visual_indexes = vision_config.deepstack_visual_indexes self.patch_embed = Qwen3_VisionPatchEmbed( patch_size=self.patch_size, temporal_patch_size=self.temporal_patch_size, in_channels=vision_config.in_channels, hidden_size=self.hidden_size, ) # vit pos embedding, TODO: spatial_patch_size vs patch_size if self.apply_vit_abs_pos_embed: self.pos_embed = nn.Embedding(self.num_grid_per_side**2, self.hidden_size) else: self.pos_embed = nn.Parameter( torch.empty([1, self.num_grid_per_side**2, self.hidden_size]) ) norm_layer = partial(nn.LayerNorm, eps=norm_eps) head_dim = self.hidden_size // self.num_heads self.rotary_pos_emb = get_rope( head_size=head_dim, max_position=8192, is_neox_style=True, rope_parameters={"partial_rotary_factor": 0.5}, ) self.blocks = nn.ModuleList( [ Qwen3_VisionBlock( dim=self.hidden_size, num_heads=self.num_heads, mlp_hidden_dim=vision_config.intermediate_size, act_fn=_ACTIVATION_REGISTRY[vision_config.hidden_act], norm_layer=norm_layer, quant_config=quant_config, prefix=f"{prefix}.blocks.{layer_idx}", ) for layer_idx in range(vision_config.depth) ] ) self.merger = Qwen3_VisionPatchMerger( d_model=vision_config.out_hidden_size, context_dim=self.hidden_size, norm_layer=norm_layer, spatial_merge_size=self.spatial_merge_size, quant_config=quant_config, prefix=f"{prefix}.merger", ) if self.deepstack_visual_indexes is not None: self.merger_list = nn.ModuleList( [ Qwen3_VisionPatchMerger( d_model=vision_config.out_hidden_size, context_dim=self.hidden_size, spatial_merge_size=self.spatial_merge_size, use_postshuffle_norm=True, norm_layer=norm_layer, quant_config=quant_config, prefix=f"{prefix}.merger_list.{layer_idx}", ) for layer_idx in range(len(self.deepstack_visual_indexes)) ] ) self.attn_backend = get_vit_attn_backend( head_size=head_dim, dtype=torch.get_default_dtype(), ) @property def dtype(self) -> torch.dtype: return self.patch_embed.proj.weight.dtype @property def device(self) -> torch.device: return self.patch_embed.proj.weight.device def rot_pos_emb(self, grid_thw): pos_ids = [] for t, h, w in grid_thw: hpos_ids = torch.arange(h).unsqueeze(1).expand(-1, w) hpos_ids = hpos_ids.reshape( h // self.spatial_merge_size, self.spatial_merge_size, w // self.spatial_merge_size, self.spatial_merge_size, ) hpos_ids = hpos_ids.permute(0, 2, 1, 3) hpos_ids = hpos_ids.flatten() wpos_ids = torch.arange(w).unsqueeze(0).expand(h, -1) wpos_ids = wpos_ids.reshape( h // self.spatial_merge_size, self.spatial_merge_size, w // self.spatial_merge_size, self.spatial_merge_size, ) wpos_ids = wpos_ids.permute(0, 2, 1, 3) wpos_ids = wpos_ids.flatten() pos_ids.append(torch.stack([hpos_ids, wpos_ids], dim=-1).repeat(t, 1)) pos_ids = torch.cat(pos_ids, dim=0) max_grid_size = grid_thw[:, 1:].max() # Use pre-computed cos_sin_cache from RotaryEmbedding cos, sin = self.rotary_pos_emb.get_cos_sin(max_grid_size) cos_combined = cos[pos_ids].flatten(1) sin_combined = sin[pos_ids].flatten(1) return cos_combined, sin_combined def fast_pos_embed_interpolate(self, grid_thw: list[list[int]]) -> torch.Tensor: num_grid_per_side = self.num_grid_per_side m_size = self.spatial_merge_size hidden_dim = self.pos_embed.embedding_dim outputs = [] for t, h, w in grid_thw: h_idxs = torch.linspace( 0, num_grid_per_side - 1, h, dtype=torch.float32, device=self.device ) w_idxs = torch.linspace( 0, num_grid_per_side - 1, w, dtype=torch.float32, device=self.device ) h_floor = h_idxs.to(torch.long) w_floor = w_idxs.to(torch.long) h_ceil = torch.clamp(h_floor + 1, max=num_grid_per_side - 1) w_ceil = torch.clamp(w_floor + 1, max=num_grid_per_side - 1) dh = h_idxs - h_floor dw = w_idxs - w_floor # Create meshgrid view for all h, w vars dh_grid, dw_grid = torch.meshgrid(dh, dw, indexing="ij") h_floor_grid, w_floor_grid = torch.meshgrid(h_floor, w_floor, indexing="ij") h_ceil_grid, w_ceil_grid = torch.meshgrid(h_ceil, w_ceil, indexing="ij") h_floor_grid_idx = h_floor_grid * num_grid_per_side h_ceil_grid_idx = h_ceil_grid * num_grid_per_side # original computation of weights # w00 = (1 - dh_grid) * (1 - dw_grid) # w01 = (1 - dh_grid) * dw_grid # w10 = dh_grid * (1 - dw_grid) # w11 = dh_grid * dw_grid # we reuse w11 here to avoid duplicate # dh_grid * dw_grid computation w11 = dh_grid * dw_grid w10 = dh_grid - w11 w01 = dw_grid - w11 w00 = 1 - dh_grid - dw_grid + w11 idx00 = h_floor_grid_idx + w_floor_grid idx01 = h_floor_grid_idx + w_ceil_grid idx10 = h_ceil_grid_idx + w_floor_grid idx11 = h_ceil_grid_idx + w_ceil_grid indices = torch.stack([idx00, idx01, idx10, idx11], dim=0).reshape(4, -1) weights = torch.stack([w00, w01, w10, w11], dim=0).reshape(4, -1, 1) weights = weights.to(dtype=self.dtype, device=self.device) embeds = self.pos_embed(indices) weighted_embeds = embeds * weights p0, p1, p2, p3 = weighted_embeds.unbind(dim=0) combined = p0 + p1 + p2 + p3 combined = combined.view(h * w, hidden_dim) repeated = combined.unsqueeze(0).expand(t, -1, -1).contiguous() repeated = repeated.view( t, h // m_size, m_size, w // m_size, m_size, hidden_dim ) repeated = repeated.permute(0, 1, 3, 2, 4, 5).reshape(-1, hidden_dim) outputs.append(repeated) return torch.cat(outputs, dim=0) def compute_attn_mask_seqlen( self, cu_seqlens: torch.Tensor, ) -> torch.Tensor: max_seqlen = torch.zeros([], device=cu_seqlens.device) if self.attn_backend in { AttentionBackendEnum.FLASH_ATTN, AttentionBackendEnum.ROCM_AITER_FA, AttentionBackendEnum.TRITON_ATTN, }: max_seqlen = (cu_seqlens[1:] - cu_seqlens[:-1]).max() return max_seqlen def forward( self, x: torch.Tensor, grid_thw: list[list[int]], ) -> torch.Tensor: hidden_states = x.to(device=self.device, dtype=self.dtype) hidden_states = self.patch_embed(hidden_states) if self.apply_vit_abs_pos_embed: pos_embeds = self.fast_pos_embed_interpolate(grid_thw) hidden_states = hidden_states + pos_embeds rotary_pos_emb_cos, rotary_pos_emb_sin = self.rot_pos_emb(grid_thw) # RDNA3 (gfx11) specific bug workaround: torch.repeat_interleave triggers # kernel crashes. We attempt the operation and catch the RuntimeError # to switch to a vectorized cumsum + searchsorted approach. try: cu_seqlens = torch.repeat_interleave( grid_thw[:, 1] * grid_thw[:, 2], grid_thw[:, 0] ).cumsum( dim=0, dtype=grid_thw.dtype if torch.jit.is_tracing() else torch.int32, ) cu_seqlens = F.pad(cu_seqlens, (1, 0), value=0) except RuntimeError: logger.warning( "torch.repeat_interleave not executable, " "switching to vectorized searchsorted implementation." ) repeat_counts = grid_thw[:, 0] values = grid_thw[:, 1] * grid_thw[:, 2] repeat_cumsum = repeat_counts.cumsum(0) total_items = repeat_cumsum[-1].item() indices = torch.searchsorted( repeat_cumsum, torch.arange(total_items, device=grid_thw.device), right=True, ) cu_seqlens = values[indices].cumsum( dim=0, dtype=grid_thw.dtype if torch.jit.is_tracing() else torch.int32, ) cu_seqlens = F.pad(cu_seqlens, (1, 0), value=0) hidden_states = hidden_states.unsqueeze(1) rotary_pos_emb_cos = rotary_pos_emb_cos.to(hidden_states.device) rotary_pos_emb_sin = rotary_pos_emb_sin.to(hidden_states.device) max_seqlen = self.compute_attn_mask_seqlen(cu_seqlens) hidden_states_list = [] deepstack_visual_indexes = self.deepstack_visual_indexes for layer_num, blk in enumerate(self.blocks): hidden_states = blk( hidden_states, cu_seqlens=cu_seqlens, rotary_pos_emb_cos=rotary_pos_emb_cos, rotary_pos_emb_sin=rotary_pos_emb_sin, max_seqlen=max_seqlen, ) if ( deepstack_visual_indexes is not None and layer_num in deepstack_visual_indexes ): hidden_states_list.append(hidden_states) hidden_states = self.merger(hidden_states) # processing deepstack if deepstack_visual_indexes is not None: processed_hidden_states_list = [hidden_states] for idx, x in enumerate(hidden_states_list): x = self.merger_list[idx](x) processed_hidden_states_list.append(x) # we cat the original visual features and deepstack features # along the feature dim hidden_states = torch.cat( processed_hidden_states_list, dim=1 ) # [seq_len, hidden_size * (1 + depth_of_deepstack)] return hidden_states def load_weights(self, weights: Iterable[tuple[str, torch.Tensor]]) -> set[str]: stacked_params_mapping = [ # (param_name, shard_name, shard_id) ("attn.qkv.", "attn.q.", "q"), ("attn.qkv.", "attn.k.", "k"), ("attn.qkv.", "attn.v.", "v"), ] params_dict = dict(self.named_parameters(remove_duplicate=False)) loaded_params: set[str] = set() for name, loaded_weight in weights: for param_name, weight_name, shard_id in stacked_params_mapping: if weight_name not in name: continue name = name.replace(weight_name, param_name) param = params_dict[name] weight_loader = param.weight_loader weight_loader(param, loaded_weight, shard_id) break else: param = params_dict[name] weight_loader = getattr(param, "weight_loader", default_weight_loader) weight_loader(param, loaded_weight) loaded_params.add(name) return loaded_params @support_torch_compile( dynamic_arg_dims={ "input_ids": 0, "positions": -1, "intermediate_tensors": 0, "inputs_embeds": 0, "deepstack_input_embeds": 0, } ) class Qwen3MoeLLMModel(Qwen3MoeModel): def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""): super().__init__(vllm_config=vllm_config, prefix=prefix) self.deepstack_multiscale_layer_start = 1 def forward( self, input_ids: torch.Tensor | None, positions: torch.Tensor, intermediate_tensors: IntermediateTensors | None = None, inputs_embeds: torch.Tensor | None = None, deepstack_input_embeds: IntermediateTensors | None = None, ) -> torch.Tensor | IntermediateTensors: if get_pp_group().is_first_rank: if inputs_embeds is not None: hidden_states = inputs_embeds else: hidden_states = self.embed_input_ids(input_ids) residual = None else: assert intermediate_tensors is not None hidden_states = intermediate_tensors["hidden_states"] residual = intermediate_tensors["residual"] for layer_idx, layer in enumerate( self.layers[self.start_layer : self.end_layer] ): layer_idx = layer_idx + self.start_layer hidden_states, residual = layer( positions, hidden_states, residual, ) if deepstack_input_embeds is not None and layer_idx in range( 0, len(deepstack_input_embeds) ): hidden_states = ( hidden_states + deepstack_input_embeds[f"deepstack_input_embeds_{layer_idx}"] ) if not get_pp_group().is_last_rank: return IntermediateTensors( {"hidden_states": hidden_states, "residual": residual} ) hidden_states, _ = self.norm(hidden_states, residual) return hidden_states class Qwen3MoeLLMForCausalLM(Qwen3MoeForCausalLM): def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""): super(Qwen3MoeForCausalLM, self).__init__() config = vllm_config.model_config.hf_config quant_config = vllm_config.quant_config self.config = config self.quant_config = quant_config self.model = Qwen3MoeLLMModel( vllm_config=vllm_config, prefix=maybe_prefix(prefix, "model") ) self.lm_head = ParallelLMHead( config.vocab_size, config.hidden_size, quant_config=quant_config ) if self.config.tie_word_embeddings: self.lm_head.weight = self.model.embed_tokens.weight self.logits_processor = LogitsProcessor(config.vocab_size) self.make_empty_intermediate_tensors = ( self.model.make_empty_intermediate_tensors ) class Qwen3OmniMoeThinkerProcessingInfo( Qwen2AudioProcessingInfo, Qwen2_5_VLProcessingInfo ): def get_hf_config(self): return self.ctx.get_hf_config(Qwen3OmniMoeConfig).thinker_config def get_hf_processor(self, **kwargs: object) -> Qwen3OmniMoeProcessor: processor = self.ctx.get_hf_processor( Qwen3OmniMoeProcessor, use_fast=kwargs.pop("use_fast", True), **kwargs, ) if not hasattr(processor, "audio_token"): processor.audio_token = "<|audio_pad|>" if not hasattr(processor, "image_token"): processor.image_token = "<|image_pad|>" if not hasattr(processor, "video_token"): processor.video_token = "<|video_pad|>" return processor def get_feature_extractor(self, **kwargs: object): hf_processor = self.get_hf_processor(**kwargs) feature_extractor = hf_processor.feature_extractor # type: ignore assert isinstance(feature_extractor, WhisperFeatureExtractor) return feature_extractor def get_supported_mm_limits(self) -> Mapping[str, int | None]: return {"audio": None, "image": None, "video": None} Qwen3OmniMoeThinkerDummyInputsBuilder = Qwen2_5OmniThinkerDummyInputsBuilder class Qwen3OmniMoeThinkerMultiModalProcessor( Qwen2_5OmniThinkerMultiModalProcessor, ): def _call_hf_processor( self, prompt: str, mm_data: Mapping[str, object], mm_kwargs: Mapping[str, object], tok_kwargs: Mapping[str, object], ) -> BatchFeature: mm_data = dict(mm_data) audios = mm_data.pop("audios", []) def pad_to_hop_length(x: np.ndarray, hop_length: int) -> np.ndarray: length = x.shape[-1] if length % hop_length != 0: pad_length = hop_length - (length % hop_length) x = np.pad(x, (0, pad_length), mode="constant", constant_values=0) return x # NOTE: WhisperFeatureExtractor cannot handle empty list of audios feature_extractor = self.info.get_feature_extractor(**mm_kwargs) hop_length = feature_extractor.hop_length if audios: # NOTE: Qwen3-Omni processor accept "audio" # To make sure the cache works with padding=True, we pre-padded # the audio to multiple of hop_length. mm_data["audio"] = [ pad_to_hop_length(audio, hop_length) if isinstance(audio, np.ndarray) else (pad_to_hop_length(audio[0], hop_length), audio[1]) for audio in audios ] # TODO(Isotr0py): Remove this patch after upstream fix PR # released and Transformers version update: # https://github.com/huggingface/transformers/pull/41473 mm_kwargs = dict(mm_kwargs) tok_kwargs = dict(tok_kwargs) mm_kwargs["audio_kwargs"] = dict(mm_kwargs.get("audio_kwargs") or {}) mm_kwargs["text_kwargs"] = dict(mm_kwargs.get("text_kwargs") or {}) if Version(TRANSFORMERS_VERSION) < Version("4.58.0"): # Extract audio_sample_rate before restructuring audio_sample_rate = mm_kwargs.pop("audio_sample_rate", None) # move truncation to audio_kwargs level to avoid conflict # with tok_kwargs mm_kwargs["audio_kwargs"].setdefault( "truncation", mm_kwargs.pop("truncation", False) ) mm_kwargs["text_kwargs"].setdefault( "truncation", tok_kwargs.pop("truncation", False) ) # Validate and conditionally pass audio_sample_rate # WhisperFeatureExtractor has a fixed sampling rate, and vLLM's # audio loader already resamples audio to the target rate. # Only pass the value if it matches to avoid unexpected behavior. if audio_sample_rate is not None: expected_sr = feature_extractor.sampling_rate if audio_sample_rate != expected_sr: logger.warning( "[%s] audio_sample_rate mismatch: user provided %dHz " "but model expects %dHz. Ignoring user value. " "vLLM's audio loader already resampled to %dHz.", self.__class__.__name__, audio_sample_rate, expected_sr, expected_sr, ) else: # Sample rate matches, safe to pass mm_kwargs["audio_kwargs"]["audio_sample_rate"] = ( audio_sample_rate ) hf_inputs = super()._call_hf_processor( prompt=prompt, mm_data=mm_data, mm_kwargs=mm_kwargs, tok_kwargs=tok_kwargs, ) if ( "audio_feature_lengths" in hf_inputs and "feature_attention_mask" in hf_inputs and (audios := mm_data.get("audio", [])) ): audio_num_frames = [] for _, audio in enumerate(audios): audio_length = len(audio[0]) if isinstance(audio, tuple) else len(audio) num_frame = ( (audio_length // hop_length) if audio_length % hop_length == 0 else (audio_length // hop_length - 1) ) if mm_kwargs.get("truncation", False): num_frame = min( num_frame, feature_extractor.n_samples // hop_length ) audio_num_frames.append(num_frame) hf_inputs["feature_attention_mask"] = [ torch.ones(num_frame) for num_frame in audio_num_frames ] hf_inputs["audio_feature_lengths"] = torch.tensor(audio_num_frames) return hf_inputs def _maybe_apply_prompt_updates( self, mm_items: MultiModalDataItems, prompt_ids: list[int], mm_kwargs: MultiModalKwargsItems, mm_prompt_updates: MultiModalPromptUpdates, is_update_applied: bool, ) -> tuple[list[int], str, Mapping[str, list[PlaceholderFeaturesInfo]]]: """ Qwen3-Omni reimplements this function to handle `use_audio_in_video`. """ mm_item_counts = mm_items.get_all_counts() self._validate_mm_kwargs(mm_kwargs, mm_item_counts) use_audio_in_video = False if "video" in mm_kwargs: for item in mm_kwargs["video"]: if item and item["use_audio_in_video"].data: use_audio_in_video = True else: use_audio_in_video = False # normal case with `use_audio_in_video=False` if is_update_applied: mm_placeholders = self._find_mm_placeholders( prompt_ids, mm_prompt_updates, ) self._validate_mm_placeholders( mm_placeholders, mm_item_counts, ) else: if use_audio_in_video and "audio" in mm_prompt_updates: filtered_updates = { k: v for k, v in mm_prompt_updates.items() if k != "audio" } prompt_ids, mm_placeholders = self._apply_prompt_updates( prompt_ids, filtered_updates, ) # Derive audio placeholders from video placeholders mm_placeholders = self._derive_audio_from_video_placeholders( mm_placeholders, mm_prompt_updates ) else: prompt_ids, mm_placeholders = self._apply_prompt_updates( prompt_ids, mm_prompt_updates, ) self._validate_mm_placeholders( mm_placeholders, mm_item_counts, ) return prompt_ids, mm_placeholders def get_updates_use_audio_in_video( self, thinker_config: PretrainedConfig, audio_len: int, video_grid_thw: list[int] | torch.Tensor, video_second_per_grid_t: float, ) -> list[int]: shift = 0 audio_token_id = thinker_config.audio_token_id video_token_id = thinker_config.video_token_id audio_start_token_id = thinker_config.audio_start_token_id audio_end_token_id = thinker_config.audio_end_token_id spatial_merge_size = thinker_config.vision_config.spatial_merge_size position_id_per_seconds = thinker_config.position_id_per_seconds audio_token_indices = np.arange(next(iter([audio_len]))) curr_video_grid_thw = next(iter([video_grid_thw])) height = curr_video_grid_thw[1] // spatial_merge_size width = curr_video_grid_thw[2] // spatial_merge_size video_token_indices = np.arange(curr_video_grid_thw[0]).reshape(-1, 1, 1) video_token_indices = np.broadcast_to( video_token_indices, (video_token_indices.shape[0], height, width) ).reshape(-1) video_token_indices = ( (video_token_indices + shift) * next(iter([video_second_per_grid_t])) * position_id_per_seconds ) video_data_index, audio_data_index = 0, 0 updates = [audio_start_token_id] while video_data_index < len(video_token_indices) and audio_data_index < len( audio_token_indices ): if ( video_token_indices[video_data_index] <= audio_token_indices[audio_data_index] ): updates += [video_token_id] video_data_index += 1 else: updates += [audio_token_id] audio_data_index += 1 if video_data_index < len(video_token_indices): updates += [video_token_id] * (len(video_token_indices) - video_data_index) if audio_data_index < len(audio_token_indices): updates += [audio_token_id] * (len(audio_token_indices) - audio_data_index) updates += [audio_end_token_id] return updates def _get_prompt_updates( self, mm_items: MultiModalDataItems, hf_processor_mm_kwargs: Mapping[str, Any], out_mm_kwargs: MultiModalKwargsItems, ) -> Sequence[PromptUpdate]: processor = self.info.get_hf_processor(**hf_processor_mm_kwargs) tokenizer = self.info.get_tokenizer() image_processor = self.info.get_image_processor(**hf_processor_mm_kwargs) vocab = tokenizer.get_vocab() audio_token = processor.audio_token image_token = processor.image_token video_token = processor.video_token audio_token_id = vocab[audio_token] image_token_id = vocab[image_token] video_token_id = vocab[video_token] out_mm_data = out_mm_kwargs.get_data() audio_feature_lengths = out_mm_data.get("audio_feature_lengths") feature_attention_mask = out_mm_data.get("feature_attention_mask") if audio_feature_lengths is None and feature_attention_mask is None: audio_output_lengths = [] elif audio_feature_lengths is not None: audio_output_lens = _get_feat_extract_output_lengths(audio_feature_lengths) audio_output_lengths = audio_output_lens.tolist() elif feature_attention_mask is not None: assert isinstance(feature_attention_mask, torch.Tensor) audio_output_lens = _get_feat_extract_output_lengths( feature_attention_mask.sum(-1) ) audio_output_lengths = audio_output_lens.tolist() # number of audios read from video. audio_in_video_item_idx = 0 audio_item_idx = 0 def get_replacement_qwen2_audio(item_idx: int): nonlocal audio_item_idx item_idx += audio_in_video_item_idx audio_item_idx += 1 num_features = audio_output_lengths[item_idx] if num_features == 0: audios = mm_items.get_items("audio", AudioProcessorItems) audio = audios.get(item_idx) raise ValueError( f"The audio {audio} (len={len(audio)}) is too short " "to be represented inside the model" ) return [audio_token_id] * num_features def get_replacement_qwen2_vision(item_idx: int, modality: str): grid_thw = out_mm_data[f"{modality}_grid_thw"][item_idx] assert isinstance(grid_thw, torch.Tensor) merge_length = image_processor.merge_size**2 token_id = image_token_id if modality == "image" else video_token_id return [token_id] * (int(grid_thw.prod()) // merge_length) use_audio_in_video = hf_processor_mm_kwargs.get("use_audio_in_video", False) thinker_config = self.info.get_hf_config() def get_replacement_qwen2_use_audio_in_video(item_idx: int): nonlocal audio_in_video_item_idx audio_num_features = audio_output_lengths[ audio_in_video_item_idx + item_idx ] video_grid_thw = out_mm_data["video_grid_thw"][item_idx] audio_in_video_item_idx += 1 second_per_grid_ts = hf_processor_mm_kwargs.get("second_per_grid_ts", None) if second_per_grid_ts: video_second_per_grid_t = second_per_grid_ts[item_idx] else: video_second_per_grid_t = 2.0 placeholder = self.get_updates_use_audio_in_video( thinker_config=thinker_config, audio_len=audio_num_features, video_grid_thw=video_grid_thw, video_second_per_grid_t=video_second_per_grid_t, ) return PromptUpdateDetails.select_token_id( placeholder, embed_token_id=video_token_id ) video_replacement_fn = ( get_replacement_qwen2_use_audio_in_video if use_audio_in_video else partial(get_replacement_qwen2_vision, modality="video") ) return [ PromptReplacement( modality="audio", target=audio_token, replacement=get_replacement_qwen2_audio, ), PromptReplacement( modality="image", target=image_token, replacement=partial(get_replacement_qwen2_vision, modality="image"), ), PromptReplacement( modality="video", target=video_token, replacement=video_replacement_fn, ), ] def _derive_audio_from_video_placeholders( self, placeholders: Mapping[str, list[PlaceholderFeaturesInfo]], mm_prompt_updates: MultiModalPromptUpdates, ) -> Mapping[str, list[PlaceholderFeaturesInfo]]: """ Helper to derive audio placeholders from video placeholders when use_audio_in_video=True. """ if "video" not in placeholders: return placeholders # Validate audio and video counts match num_videos = len(placeholders["video"]) num_audios = len(mm_prompt_updates.get("audio", [])) if num_audios != num_videos: raise ValueError( f"use_audio_in_video requires equal number of audio and video items, " f"got {num_audios=}, {num_videos=}" ) tokenizer = self.info.get_tokenizer() processor = self.info.get_hf_processor() audio_token_id = tokenizer.get_vocab()[processor.audio_token] result_placeholders = dict(placeholders) audio_placeholders = [] # Each video is paired with one audio for video_idx, video_placeholder in enumerate(placeholders["video"]): # Create is_embed mask selecting only audio tokens audio_is_embed = torch.tensor(video_placeholder.tokens) == audio_token_id audio_placeholder = PlaceholderFeaturesInfo( modality="audio", item_idx=video_idx, start_idx=video_placeholder.start_idx, tokens=video_placeholder.tokens, is_embed=audio_is_embed, ) audio_placeholders.append(audio_placeholder) result_placeholders["audio"] = audio_placeholders return result_placeholders def _get_raw_input_ids( self, token_ids: list[int], use_audio_in_video: bool = False, ) -> list[int]: tokenizer = self.info.get_tokenizer() vision_bos_token = tokenizer.encode(tokenizer.vision_bos_token)[0] vision_eos_token = tokenizer.encode(tokenizer.vision_eos_token)[0] audio_bos_token = tokenizer.encode(tokenizer.audio_bos_token)[0] audio_eos_token = tokenizer.encode(tokenizer.audio_eos_token)[0] audio_token = tokenizer.encode("<|audio_pad|>")[0] image_token = tokenizer.encode("<|image_pad|>")[0] video_token = tokenizer.encode("<|video_pad|>")[0] result = token_ids[:] if use_audio_in_video: while True: start = None for i in range(len(result) - 1): if result[i : i + 2] == [vision_bos_token, audio_bos_token]: start = i break if start is not None: end = None for i in range(start + 2, len(result) - 1): if result[i : i + 2] == [audio_eos_token, vision_eos_token]: end = i break if end is not None: result = ( result[:start] + [vision_bos_token, video_token, vision_eos_token] + result[end + 2 :] ) else: break for mm_token in [audio_token, image_token, video_token]: compressed = [] for x in result: if x != mm_token or (not compressed or compressed[-1] != mm_token): compressed.append(x) result = compressed return result class Qwen3OmniMoeConditionalGenerationMixin(Qwen2_5OmniConditionalGenerationMixin): def _process_audio_input( self, audio_input: Qwen2_5OmniAudioFeatureInputs, audio_hashes: list[str] | None = None, cached_audio_features: torch.Tensor | None = None, ) -> tuple[torch.Tensor, ...]: input_features = audio_input["input_features"] audio_feature_lengths = audio_input["audio_feature_lengths"] audio_output_lengths = _get_feat_extract_output_lengths(audio_feature_lengths) audio_features = self.audio_tower( input_features.to(self.audio_tower.dtype), feature_lens=audio_feature_lengths, aftercnn_lens=audio_output_lengths, ) return audio_features.split(audio_output_lengths.tolist()) @MULTIMODAL_REGISTRY.register_processor( Qwen3OmniMoeThinkerMultiModalProcessor, info=Qwen3OmniMoeThinkerProcessingInfo, dummy_inputs=Qwen3OmniMoeThinkerDummyInputsBuilder, ) class Qwen3OmniMoeThinkerForConditionalGeneration( nn.Module, SupportsMultiModal, SupportsPP, SupportsMRoPE, Qwen3OmniMoeConditionalGenerationMixin, SupportsTranscription, ): hf_to_vllm_mapper = WeightsMapper( orig_to_new_prefix={ "thinker.lm_head.": "language_model.lm_head.", "thinker.model.": "language_model.model.", "thinker.": "", } ) packed_modules_mapping = { "qkv_proj": [ "q_proj", "k_proj", "v_proj", ], "gate_up_proj": [ "gate_proj", "up_proj", ], } supported_languages = ISO639_1_SUPPORTED_LANGS @classmethod def get_placeholder_str(cls, modality: str, i: int) -> str | None: if modality.startswith("image"): return "<|vision_start|><|image_pad|><|vision_end|>" if modality.startswith("video"): return "<|vision_start|><|video_pad|><|vision_end|>" if modality.startswith("audio"): return "<|audio_start|><|audio_pad|><|audio_end|>" raise ValueError("Only image, video or audio modality is supported") def __init__(self, *, vllm_config: VllmConfig, prefix: str = ""): super().__init__() self.vllm_config = vllm_config # needed for torch compile forward context thinker_config: Qwen3OmniMoeThinkerConfig = ( vllm_config.model_config.hf_config.thinker_config ) quant_config = vllm_config.quant_config multimodal_config = vllm_config.model_config.multimodal_config self.config = thinker_config self.multimodal_config = multimodal_config self.quant_config = quant_config with self._mark_tower_model(vllm_config, "audio"): self.audio_tower = Qwen3OmniMoeAudioEncoder( thinker_config.audio_config, prefix=maybe_prefix(prefix, "audio_tower"), ) self.use_deepstack = hasattr( thinker_config.vision_config, "deepstack_visual_indexes" ) self.deepstack_num_level = ( len(thinker_config.vision_config.deepstack_visual_indexes) if self.use_deepstack else 0 ) self.visual_dim = thinker_config.vision_config.out_hidden_size self.multiscale_dim = self.visual_dim * self.deepstack_num_level with self._mark_tower_model(vllm_config, {"image", "video"}): self.visual = Qwen3Omni_VisionTransformer( vision_config=thinker_config.vision_config, norm_eps=getattr(thinker_config.text_config, "rms_norm_eps", 1e-6), quant_config=quant_config, prefix=maybe_prefix(prefix, "visual"), ) # register buffer for deepstack if self.use_deepstack: self.deepstack_input_embeds = [ torch.zeros( vllm_config.scheduler_config.max_num_batched_tokens, thinker_config.text_config.hidden_size, ) for _ in range(self.deepstack_num_level) ] with self._mark_language_model(vllm_config): self.language_model = Qwen3MoeLLMForCausalLM( vllm_config=vllm_config.with_hf_config( thinker_config.text_config, architectures=["Qwen3MoeForCausalLM"], ), prefix=maybe_prefix(prefix, "language_model"), ) self.make_empty_intermediate_tensors = ( self.language_model.make_empty_intermediate_tensors ) def _get_deepstack_input_embeds( self, num_tokens: int, ) -> IntermediateTensors | None: if not getattr(self, "deepstack_input_embeds", None): return None # If vision tower is skipped # get deepstack_input_embeds from buffer, and clear the buffer return IntermediateTensors( { f"deepstack_input_embeds_{idx}": self.deepstack_input_embeds[idx][ :num_tokens ] for idx in range(self.deepstack_num_level) } ) def _set_deepstack_input_embeds(self, deepstack_input_embeds: torch.Tensor) -> None: if not getattr(self, "deepstack_input_embeds", None): return # set deepstack_input_embeds to buffer num_tokens = deepstack_input_embeds.size(1) if num_tokens > self.deepstack_input_embeds[0].size(0): self.deepstack_input_embeds = [ torch.zeros( num_tokens, self.config.text_config.hidden_size, device=self.deepstack_input_embeds[0].device, dtype=self.deepstack_input_embeds[0].dtype, ) for _ in range(self.deepstack_num_level) ] for idx in range(self.deepstack_num_level): self.deepstack_input_embeds[idx][:num_tokens].copy_( deepstack_input_embeds[idx] ) def _clear_deepstack_input_embeds(self, num_tokens: int) -> None: if not getattr(self, "deepstack_input_embeds", None): return # clear deepstack_input_embeds in buffer if num_tokens > 0: for idx in range(self.deepstack_num_level): self.deepstack_input_embeds[idx][:num_tokens].zero_() def _parse_and_validate_multimodal_inputs(self, **kwargs: object) -> dict: mm_input_by_modality = {} # Preserve the order of modalities if there are multiple of them # from the order of kwargs. for input_key in kwargs: if ( input_key in ("pixel_values", "image_embeds") and "image" not in mm_input_by_modality ): mm_input_by_modality["image"] = self._parse_and_validate_image_input( **kwargs ) if ( input_key in ("pixel_values_videos", "video_embeds") and "video" not in mm_input_by_modality ): mm_input_by_modality["video"] = self._parse_and_validate_video_input( **kwargs ) if ( input_key in ("input_audio_features") and "audio" not in mm_input_by_modality ): mm_input_by_modality["audio"] = self._parse_and_validate_audio_input( **kwargs ) return mm_input_by_modality def embed_multimodal(self, **kwargs: object) -> MultiModalEmbeddings | None: mm_input_by_modality = self._parse_and_validate_multimodal_inputs(**kwargs) if not mm_input_by_modality: return [] # The result multimodal_embeddings is tuple of tensors, with each # tensor correspoending to a multimodal data item (image or video). multimodal_embeddings: tuple[torch.Tensor, ...] = () # NOTE: It is important to iterate over the keys in this dictionary # to preserve the order of the modalities. for modality in mm_input_by_modality: multimodal_input = mm_input_by_modality[modality] if modality == "image": image_embeddings = self._process_image_input(multimodal_input) multimodal_embeddings += tuple(image_embeddings) if modality == "video": video_embeddings = self._process_video_input(multimodal_input) multimodal_embeddings += tuple(video_embeddings) if modality == "audio": audio_embeddings = self._process_audio_input(multimodal_input) multimodal_embeddings += tuple(audio_embeddings) return multimodal_embeddings def embed_input_ids( self, input_ids: torch.Tensor, multimodal_embeddings: MultiModalEmbeddings | None = None, *, is_multimodal: torch.Tensor | None = None, handle_oov_mm_token: bool = False, ) -> torch.Tensor: inputs_embeds = self._embed_text_input_ids( input_ids, self.language_model.embed_input_ids, is_multimodal=is_multimodal, handle_oov_mm_token=handle_oov_mm_token, ) if multimodal_embeddings is None or len(multimodal_embeddings) == 0: return inputs_embeds # Detect interleaved audio-in-video early, since it affects # both the deepstack path and the final embedding merge. video_token_id = self.config.video_token_id audio_token_id = self.config.audio_token_id is_video = is_multimodal & (input_ids == video_token_id) is_audio = is_multimodal & (input_ids == audio_token_id) num_video = is_video.sum().item() num_audio = is_audio.sum().item() is_interleaved = check_interleaved_audio_video( is_video, is_audio, num_video, num_audio ) deepstack_input_embeds = None # split the feat dim to obtain multi-scale visual feature has_vision_embeddings = [ embeddings.shape[-1] != self.config.text_config.hidden_size for embeddings in multimodal_embeddings ] if self.visual.deepstack_visual_indexes is not None and any( has_vision_embeddings ): multiscale_len = len(self.visual.deepstack_visual_indexes) multimodal_embeddings_multiscale = [] if is_interleaved: # Use input_ids-based mask for correct vision positions # when audio and video tokens are interleaved. is_vision = is_video.clone() else: is_vision = torch.zeros_like(is_multimodal) mm_positions = torch.nonzero(is_multimodal, as_tuple=True)[0] mm_position_idx = 0 for index, embeddings in enumerate(multimodal_embeddings): num_tokens = embeddings.shape[0] # Vision embeddings if embeddings.shape[-1] != self.config.text_config.hidden_size: visual_dim = embeddings.shape[-1] // (multiscale_len + 1) multi_dim = visual_dim * multiscale_len embeddings_main, embeddings_multiscale = torch.split( embeddings, [visual_dim, multi_dim], dim=-1 ) multimodal_embeddings[index] = embeddings_main multimodal_embeddings_multiscale.append(embeddings_multiscale) if not is_interleaved: current_positions = mm_positions[ mm_position_idx : mm_position_idx + num_tokens ] is_vision[current_positions] = True # Audio embeddings else: if not is_interleaved: current_positions = mm_positions[ mm_position_idx : mm_position_idx + num_tokens ] is_vision[current_positions] = False if not is_interleaved: mm_position_idx += num_tokens deepstack_input_embeds = inputs_embeds.new_zeros( inputs_embeds.size(0), multiscale_len * inputs_embeds.size(1) ) deepstack_input_embeds = _merge_multimodal_embeddings( inputs_embeds=deepstack_input_embeds, multimodal_embeddings=multimodal_embeddings_multiscale, is_multimodal=is_vision, ) deepstack_input_embeds = ( deepstack_input_embeds.view( inputs_embeds.shape[0], multiscale_len, visual_dim ) .permute(1, 0, 2) .contiguous() ) self._set_deepstack_input_embeds(deepstack_input_embeds) if is_interleaved: return merge_interleaved_embeddings( inputs_embeds, multimodal_embeddings, is_video, is_audio, is_multimodal, num_video, num_audio, ) # Default: standard merge (no interleaving) inputs_embeds = _merge_multimodal_embeddings( inputs_embeds=inputs_embeds, multimodal_embeddings=multimodal_embeddings, is_multimodal=is_multimodal, ) return inputs_embeds def forward( self, input_ids: torch.Tensor | None, positions: torch.Tensor, intermediate_tensors: IntermediateTensors | None = None, inputs_embeds: torch.Tensor | None = None, **kwargs: object, ) -> torch.Tensor | IntermediateTensors: if intermediate_tensors is not None: inputs_embeds = None if inputs_embeds is not None and get_pp_group().is_first_rank: deepstack_input_embeds = self._get_deepstack_input_embeds( inputs_embeds.size(0) ) else: deepstack_input_embeds = None hidden_states = self.language_model.model( input_ids, positions, intermediate_tensors, inputs_embeds=inputs_embeds, # args for deepstack deepstack_input_embeds=deepstack_input_embeds, ) if inputs_embeds is not None and get_pp_group().is_first_rank: self._clear_deepstack_input_embeds(inputs_embeds.size(0)) return hidden_states def compute_logits( self, hidden_states: torch.Tensor, ) -> torch.Tensor | None: return self.language_model.compute_logits(hidden_states) def load_weights(self, weights: Iterable[tuple[str, torch.Tensor]]) -> set[str]: loader = AutoWeightsLoader( self, skip_prefixes=["talker.", "code2wav."], ) loaded_weights = loader.load_weights(weights, mapper=self.hf_to_vllm_mapper) return loaded_weights def _compute_audio_token_count(self, audio_feature_length: int) -> int: """Compute audio tokens from feature length using Qwen3-Omni formula.""" return _get_feat_extract_output_lengths( torch.tensor([audio_feature_length]) ).item() def _get_audio_for_video_mapping( self, mm_features: list[MultiModalFeatureSpec] ) -> tuple[dict[int, int], set[int]]: """ Map video offset -> paired audio_feature_length for use_audio_in_video. When use_audio_in_video=True, audio is interleaved within video. The pairing is based on feature order in mm_features. Returns: Tuple of (video_offset -> audio_feature_length mapping, set of paired audio offsets to skip) """ videos_with_audio = [ f for f in mm_features if f.modality == "video" and f.data.get("use_audio_in_video") and f.data["use_audio_in_video"].data.item() ] audios = [f for f in mm_features if f.modality == "audio"] mapping: dict[int, int] = {} paired_audio_offsets: set[int] = set() for i, video_f in enumerate(videos_with_audio): if i < len(audios): audio_len = audios[i].data["audio_feature_lengths"].data.item() mapping[video_f.mm_position.offset] = audio_len paired_audio_offsets.add(audios[i].mm_position.offset) return mapping, paired_audio_offsets def iter_mm_features( self, mm_features: list[MultiModalFeatureSpec] ) -> Iterator[tuple[int, str, dict[str, Any]]]: """ Iterate over multimodal features sorted by position offset. Yields: (offset, modality, feature_data) where feature_data contains: - image: {"grid_t", "grid_h", "grid_w", "t_factor"} - video: {"grid_t", "grid_h", "grid_w", "t_factor", "use_audio_in_video", "audio_feature_length"} - audio: {"audio_feature_length"} """ config = self.config spatial_merge_size = config.vision_config.spatial_merge_size position_id_per_seconds = config.position_id_per_seconds sorted_features = sorted(mm_features, key=lambda f: f.mm_position.offset) audio_for_video, paired_audio_offsets = self._get_audio_for_video_mapping( sorted_features ) for mm_feature in sorted_features: offset = mm_feature.mm_position.offset modality = mm_feature.modality if modality == "image": t, h, w = mm_feature.data["image_grid_thw"].data.tolist() yield ( offset, "image", { "grid_t": t, "grid_h": h // spatial_merge_size, "grid_w": w // spatial_merge_size, "t_factor": position_id_per_seconds, }, ) elif modality == "video": t, h, w = mm_feature.data["video_grid_thw"].data.tolist() second_per_grid_ts = 2.0 if mm_feature.data.get("second_per_grid_ts"): second_per_grid_ts = mm_feature.data[ "second_per_grid_ts" ].data.item() use_audio_in_video = bool( mm_feature.data.get("use_audio_in_video") and mm_feature.data["use_audio_in_video"].data.item() ) yield ( offset, "video", { "grid_t": t, "grid_h": h // spatial_merge_size, "grid_w": w // spatial_merge_size, "t_factor": second_per_grid_ts * position_id_per_seconds, "use_audio_in_video": use_audio_in_video, "audio_feature_length": audio_for_video.get(offset), }, ) elif modality == "audio": if offset not in paired_audio_offsets: audio_len = mm_feature.data["audio_feature_lengths"].data.item() yield offset, "audio", {"audio_feature_length": audio_len} def _compute_interleaved_positions( self, start_idx: int, data: dict[str, Any] ) -> tuple[np.ndarray, int]: """ Compute positions for interleaved video+audio using Qwen3 token-by-token interleaving logic. Returns: (position_ids [3, N], total_token_count) """ grid_t = data["grid_t"] grid_h = data["grid_h"] grid_w = data["grid_w"] t_factor = data["t_factor"] audio_feature_length = data["audio_feature_length"] audio_len = self._compute_audio_token_count(audio_feature_length) h_index = np.tile( np.arange(grid_h).reshape(1, -1, 1), (grid_t, 1, grid_w) ).flatten() w_index = np.tile( np.arange(grid_w).reshape(1, 1, -1), (grid_t, grid_h, 1) ).flatten() t_index_raw = np.arange(grid_t) t_index_scaled = (t_index_raw * t_factor).astype(np.int64) t_index = np.repeat(t_index_scaled, grid_h * grid_w) video_pos = np.stack([t_index, h_index, w_index]) + start_idx audio_pos = np.broadcast_to(np.arange(audio_len), (3, audio_len)) + start_idx video_t_values = video_pos[0] audio_t_values = audio_pos[0] pos_ids_list: list[np.ndarray] = [] video_idx, audio_idx = 0, 0 num_video = grid_t * grid_h * grid_w while video_idx < num_video and audio_idx < audio_len: if video_t_values[video_idx] <= audio_t_values[audio_idx]: pos_ids_list.append(video_pos[:, video_idx : video_idx + 1]) video_idx += 1 else: pos_ids_list.append(audio_pos[:, audio_idx : audio_idx + 1]) audio_idx += 1 if video_idx < num_video: pos_ids_list.append(video_pos[:, video_idx:]) if audio_idx < audio_len: pos_ids_list.append(audio_pos[:, audio_idx:]) total_tokens = num_video + audio_len return np.concatenate(pos_ids_list, axis=1), total_tokens @classmethod def get_speech_to_text_config( cls, model_config: ModelConfig, task_type: str ) -> SpeechToTextConfig: processor = cached_processor_from_config( model_config, processor_cls=Qwen3OmniMoeProcessor ) return SpeechToTextConfig( max_audio_clip_s=processor.feature_extractor.chunk_length, sample_rate=processor.feature_extractor.sampling_rate, min_energy_split_window_size=None, ) @classmethod def get_generation_prompt( cls, audio: np.ndarray, stt_config: SpeechToTextConfig, model_config: ModelConfig, language: str | None, task_type: Literal["transcribe", "translate"], request_prompt: str, to_language: str | None, ) -> PromptType: """ Construct a transcription/translation prompt for Qwen3-Omni. """ # Transcribe this audio [into ] | for transcription # Translate this audio [from into ] | for translation instruction = "Transcribe" if task_type == "transcribe" else "Translate" instruction += " this audio" # Default to_language to English for translation if task_type == "translate" and to_language is None: to_language = "en" # Get full language names from supported_languages mapping full_lang_name = cls.supported_languages.get(language, "") full_lang_name_to = cls.supported_languages.get(to_language, "") if task_type == "transcribe" and full_lang_name: instruction += f" into {full_lang_name}" elif task_type == "translate": if full_lang_name: instruction += f" from {full_lang_name}" if full_lang_name_to: instruction += f" into {full_lang_name_to}" instruction += "." if request_prompt: instruction += f" {request_prompt}" processor = cached_processor_from_config( model_config, processor_cls=Qwen3OmniMoeProcessor ) # Audio placeholder format: <|audio_start|><|audio_pad|><|audio_end|> audio_placeholder = "<|audio_start|><|audio_pad|><|audio_end|>" user_content = f"{audio_placeholder}{instruction}" messages = [{"role": "user", "content": user_content}] prompt = processor.apply_chat_template( messages, tokenize=False, add_generation_prompt=True, ) audio_data = (audio, stt_config.sample_rate) prompts_dict = {"multi_modal_data": {"audio": audio_data}, "prompt": prompt} return cast(PromptType, prompts_dict) def get_mrope_input_positions( self, input_tokens: list[int], mm_features: list[MultiModalFeatureSpec], ) -> tuple[torch.Tensor, int]: """Compute M-RoPE input positions using mm_features directly.""" seq_len = len(input_tokens) llm_pos_ids_list: list[np.ndarray] = [] st = 0 for offset, modality, data in self.iter_mm_features(mm_features): text_len = offset - st st_idx = int(llm_pos_ids_list[-1].max()) + 1 if llm_pos_ids_list else 0 if text_len > 0: llm_pos_ids_list.append( np.broadcast_to(np.arange(text_len), (3, text_len)) + st_idx ) st_idx += text_len bos_pos = np.broadcast_to(np.array([st_idx]), (3, 1)) llm_pos_ids_list.append(bos_pos) st_idx += 1 if modality == "audio": audio_tokens = self._compute_audio_token_count( data["audio_feature_length"] ) audio_pos = ( np.broadcast_to(np.arange(audio_tokens), (3, audio_tokens)) + st_idx ) llm_pos_ids_list.append(audio_pos) st_idx = int(audio_pos.max()) + 1 eos_pos = np.broadcast_to(np.array([st_idx]), (3, 1)) llm_pos_ids_list.append(eos_pos) st = offset + 1 + audio_tokens + 1 elif modality == "image": grid_t = data["grid_t"] grid_h = data["grid_h"] grid_w = data["grid_w"] t_factor = data["t_factor"] grid_indices = np.indices((grid_t, grid_h, grid_w)) if t_factor != 1.0: grid_indices[0] = (grid_indices[0] * t_factor).astype(np.int64) llm_pos_ids_list.append(grid_indices.reshape(3, -1) + st_idx) image_len = grid_t * grid_h * grid_w st_idx = int(llm_pos_ids_list[-1].max()) + 1 eos_pos = np.broadcast_to(np.array([st_idx]), (3, 1)) llm_pos_ids_list.append(eos_pos) st = offset + 1 + image_len + 1 elif modality == "video": grid_t = data["grid_t"] grid_h = data["grid_h"] grid_w = data["grid_w"] t_factor = data["t_factor"] if not data["use_audio_in_video"]: grid_indices = np.indices((grid_t, grid_h, grid_w)) if t_factor != 1.0: grid_indices[0] = (grid_indices[0] * t_factor).astype(np.int64) llm_pos_ids_list.append(grid_indices.reshape(3, -1) + st_idx) video_len = grid_t * grid_h * grid_w st_idx = int(llm_pos_ids_list[-1].max()) + 1 eos_pos = np.broadcast_to(np.array([st_idx]), (3, 1)) llm_pos_ids_list.append(eos_pos) st = offset + 1 + video_len + 1 else: audio_bos_pos = np.broadcast_to(np.array([st_idx - 1]), (3, 1)) llm_pos_ids_list.append(audio_bos_pos) pos_ids, _ = self._compute_interleaved_positions(st_idx, data) llm_pos_ids_list.append(pos_ids) st_idx = int(pos_ids.max()) + 1 eos_pos = np.broadcast_to(np.array([st_idx]), (3, 1)) llm_pos_ids_list.append(eos_pos) llm_pos_ids_list.append(eos_pos) video_len = grid_t * grid_h * grid_w audio_len = self._compute_audio_token_count( data["audio_feature_length"] ) st = offset + 2 + video_len + audio_len + 2 if st < seq_len: st_idx = int(llm_pos_ids_list[-1].max()) + 1 if llm_pos_ids_list else 0 text_len = seq_len - st llm_pos_ids_list.append( np.broadcast_to(np.arange(text_len), (3, text_len)) + st_idx ) llm_positions = np.concatenate(llm_pos_ids_list, axis=1).reshape(3, -1) if llm_positions.shape[1] != seq_len: raise RuntimeError("Position ids length mismatch with input ids length") mrope_position_delta = int(llm_positions.max()) + 1 - seq_len return torch.from_numpy(llm_positions), mrope_position_delta def get_mm_mapping(self) -> MultiModelKeys: """ Get the module prefix in multimodal models """ return MultiModelKeys.from_string_field( language_model="language_model", connector="visual.merger", tower_model=["visual.", "audio_tower."], )