# Qwen4-Exp

Qwen4-Exp builds on Qwen3.5's hybrid text and multimodal architecture with three key components: GatedResidual (GR), Qwen Sparse Attention (QSA), and Per-Layer Embedding (PLE).

GR is a Qwen-developed residual architecture that combines Hyper-Connection with GatedNorm. It mixes multiple residual streams with fine-grained elementwise gating before each attention and Mixture-of-Experts (MoE) block, then controls how much of the block output is injected back into each stream.

QSA uses multiple query heads to score compressed key blocks, selects the most relevant contiguous token blocks, and keeps the incomplete trailing block uncompressed. This block-level selection reduces indexing overhead and improves memory locality for long sequences. Combined with Gated DeltaNet, QSA makes Qwen4-Exp the first hybrid architecture to integrate linear and sparse attention, substantially improving inference efficiency for long-context workloads.

PLE enriches selected decoder layers with layer-specific lexical features derived from hashed token n-grams and a dilated depthwise convolution.

## Usage tips

- `ple_layer_ids` uses one-based decoder layer indices. When PLE is enabled and the model receives `inputs_embeds`, pass the original token ids through `ple_input_ids`. If `input_ids` are provided, the model uses them for PLE automatically.
- During cached generation, PLE maintains both n-gram context and dilated-convolution state. [DynamicCache](/docs/transformers/main/en/internal/generation_utils#transformers.DynamicCache) is the recommended default, and [StaticCache](/docs/transformers/main/en/internal/generation_utils#transformers.StaticCache) is also supported. Cache cropping is not supported when PLE or QSA is enabled. With cache offloading, GatedDeltaNet, PLE, and QSA indexer states remain on device while attention key/value states are offloaded.
- `split_ngram_parts` controls the logical checkpoint shards for each large PLE n-gram embedding table. Transformers concatenates these shards along the vocabulary dimension into one runtime weight. The default `save_pretrained(save_original_format=True)` writes the configured original sharded layout.
- `hc_count` sets the number of residual streams, and `hc_lowrank` sets the rank of the learned GR input mixer.
- Providing the complete set of QSA indexer fields enables sparse token selection on full-attention layers. Eager and SDPA attention are supported; other requested backends fall back to eager. Automatic generation compilation is disabled because token selection is data-dependent.
- `tp_plan="auto"` supports attention, MoE, GatedDeltaNet, QSA, GR, and vocabulary-row sharding of PLE tables. Custom configurations must keep sharded attention heads, QSA and GatedDeltaNet projection dimensions, GR stream width, expert dimensions, and the padded PLE vocabulary divisible by the TP size. The replicated low-rank GR output adds no divisibility constraint.
- FSDP2 shards token embeddings and decoder layers while keeping the final GR mixer gathered. TP and FSDP cannot currently be combined, and no pipeline-parallel plan is provided.
- Use [Qwen4ExpForCausalLM](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpForCausalLM) with [Qwen4ExpTextConfig](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpTextConfig) for text-only generation. Use [Qwen4ExpForConditionalGeneration](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpForConditionalGeneration) with [Qwen4ExpConfig](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpConfig) for multimodal inputs.

## Qwen4ExpConfig[[transformers.Qwen4ExpConfig]]

#### transformers.Qwen4ExpConfig[[transformers.Qwen4ExpConfig]]

```python
transformers.Qwen4ExpConfig(transformers_version: str | None = None, architectures: list[str] | None = None, output_hidden_states: bool | None = False, return_dict: bool | None = True, dtype: typing.Union[str, ForwardRef('torch.dtype'), NoneType] = None, chunk_size_feed_forward: int = 0, is_encoder_decoder: bool = False, id2label: dict[int, str] | dict[str, str] | None = None, label2id: dict[str, int] | dict[str, str] | None = None, problem_type: typing.Optional[typing.Literal['regression', 'single_label_classification', 'multi_label_classification']] = None, text_config: dict | transformers.configuration_utils.PreTrainedConfig | None = None, vision_config: dict | transformers.configuration_utils.PreTrainedConfig | None = None, image_token_id: int = 248056, video_token_id: int = 248057, vision_start_token_id: int = 248053, vision_end_token_id: int = 248054, tie_word_embeddings: bool = False)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/configuration_qwen4_exp.py#L290)

**Parameters:**

text_config (`Union[dict, ~configuration_utils.PreTrainedConfig]`, *optional*) : The config object or dictionary of the text backbone.

vision_config (`Union[dict, ~configuration_utils.PreTrainedConfig]`, *optional*) : The config object or dictionary of the vision backbone.

image_token_id (`int`, *optional*, defaults to `248056`) : The image token index used as a placeholder for input images.

video_token_id (`int`, *optional*, defaults to `248057`) : The video token index used as a placeholder for input videos.

vision_start_token_id (`int`, *optional*, defaults to `248053`) : Token ID that marks the start of a visual segment in the multimodal input sequence.

vision_end_token_id (`int`, *optional*, defaults to `248054`) : Token ID that marks the end of a visual segment in the multimodal input sequence.

tie_word_embeddings (`bool`, *optional*, defaults to `False`) : Whether to tie weight embeddings according to model's `tied_weights_keys` mapping.

This is the configuration class to store the configuration of a Qwen4ExpModel. It is used to instantiate a Qwen4 Exp
model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
defaults will yield a similar configuration to that of the [Qwen/Qwen4-Exp](https://huggingface.co/Qwen/Qwen4-Exp)

Configuration objects inherit from [PreTrainedConfig](/docs/transformers/main/en/main_classes/configuration#transformers.PreTrainedConfig) and can be used to control the model outputs. Read the
documentation from [PreTrainedConfig](/docs/transformers/main/en/main_classes/configuration#transformers.PreTrainedConfig) for more information.

Example:

```python
>>> from transformers import Qwen4ExpConfig, Qwen4ExpForConditionalGeneration

>>> configuration = Qwen4ExpConfig()
>>> model = Qwen4ExpForConditionalGeneration(configuration)
>>> configuration = model.config
```

## Qwen4ExpTextConfig[[transformers.Qwen4ExpTextConfig]]

#### transformers.Qwen4ExpTextConfig[[transformers.Qwen4ExpTextConfig]]

```python
transformers.Qwen4ExpTextConfig(transformers_version: str | None = None, architectures: list[str] | None = None, output_hidden_states: bool | None = False, return_dict: bool | None = True, dtype: typing.Union[str, ForwardRef('torch.dtype'), NoneType] = None, chunk_size_feed_forward: int = 0, is_encoder_decoder: bool = False, id2label: dict[int, str] | dict[str, str] | None = None, label2id: dict[str, int] | dict[str, str] | None = None, problem_type: typing.Optional[typing.Literal['regression', 'single_label_classification', 'multi_label_classification']] = None, vocab_size: int = 248320, hidden_size: int = 2048, num_hidden_layers: int = 40, num_attention_heads: int = 16, num_key_value_heads: int = 2, hidden_act: str = 'silu', max_position_embeddings: int = 32768, initializer_range: float = 0.02, rms_norm_eps: float = 1e-06, use_cache: bool = True, tie_word_embeddings: bool = False, rope_parameters: transformers.modeling_rope_utils.RopeParameters | dict | None = None, attention_bias: bool = False, attention_dropout: float | int = 0.0, head_dim: int = 256, linear_conv_kernel_dim: int = 4, linear_key_head_dim: int = 128, linear_value_head_dim: int = 128, linear_num_key_heads: int = 16, linear_num_value_heads: int = 32, moe_intermediate_size: int = 512, shared_expert_intermediate_size: int = 512, num_experts_per_tok: int = 10, num_experts: int = 512, output_router_logits: bool = False, router_aux_loss_coef: float = 0.001, layer_types: list[str] | None = None, pad_token_id: int | None = None, bos_token_id: int | None = None, eos_token_id: int | list[int] | None = None, hc_count: int = 4, hc_lowrank: int = 320, ple_layer_ids: list[int] | None = None, ple_embed_dim: int | None = None, ple_conv_kernel_size: int = 4, ngram_size: int = 3, heads_per_ngram: int = 8, ngram_vocab_size_base: int = 20000000, make_ngram_vocab_size_divisible_by: int = 128, seed: int = 1234, split_ngram_parts: int = 512, indexer_n_heads: int | None = None, indexer_kv_heads: int | None = None, indexer_head_dim: int | None = None, indexer_budget: int | None = None, indexer_compress_ratio: int | None = None, norm_topk_prob: bool = True, output_gate_type: str | None = None)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/configuration_qwen4_exp.py#L29)

**Parameters:**

vocab_size (`int`, *optional*, defaults to `248320`) : Vocabulary size of the model. Defines the number of different tokens that can be represented by the `input_ids`.

hidden_size (`int`, *optional*, defaults to `2048`) : Dimension of the hidden representations.

num_hidden_layers (`int`, *optional*, defaults to `40`) : Number of hidden layers in the Transformer decoder.

num_attention_heads (`int`, *optional*, defaults to `16`) : Number of attention heads for each attention layer in the Transformer decoder.

num_key_value_heads (`int`, *optional*, defaults to `2`) : This is the number of key_value heads that should be used to implement Grouped Query Attention. If `num_key_value_heads=num_attention_heads`, the model will use Multi Head Attention (MHA), if `num_key_value_heads=1` the model will use Multi Query Attention (MQA) otherwise GQA is used. When converting a multi-head checkpoint to a GQA checkpoint, each group key and value head should be constructed by meanpooling all the original heads within that group. For more details, check out [this paper](https://huggingface.co/papers/2305.13245). If it is not specified, will default to `num_attention_heads`.

hidden_act (`str`, *optional*, defaults to `silu`) : The non-linear activation function (function or string) in the decoder. For example, `"gelu"`, `"relu"`, `"silu"`, etc.

max_position_embeddings (`int`, *optional*, defaults to `32768`) : The maximum sequence length that this model might ever be used with.

initializer_range (`float`, *optional*, defaults to `0.02`) : The standard deviation of the truncated_normal_initializer for initializing all weight matrices.

rms_norm_eps (`float`, *optional*, defaults to `1e-06`) : The epsilon used by the rms normalization layers.

use_cache (`bool`, *optional*, defaults to `True`) : Whether or not the model should return the last key/values attentions (not used by all models). Only relevant if `config.is_decoder=True` or when the model is a decoder-only generative model.

tie_word_embeddings (`bool`, *optional*, defaults to `False`) : Whether to tie weight embeddings according to model's `tied_weights_keys` mapping.

rope_parameters (`Union[~modeling_rope_utils.RopeParameters, dict]`, *optional*) : Dictionary containing the configuration parameters for the RoPE embeddings. The dictionary should contain a value for `rope_theta` and optionally parameters used for scaling in case you want to use RoPE with longer `max_position_embeddings`.

attention_bias (`bool`, *optional*, defaults to `False`) : Whether to use a bias in the query, key, value and output projection layers during self-attention.

attention_dropout (`Union[float, int]`, *optional*, defaults to `0.0`) : The dropout ratio for the attention probabilities.

head_dim (`int`, *optional*, defaults to `256`) : The attention head dimension. If None, it will default to hidden_size // num_attention_heads

linear_conv_kernel_dim (`int`, *optional*, defaults to 4) : Kernel size of the convolution used in linear attention layers.

linear_key_head_dim (`int`, *optional*, defaults to 128) : Dimension of each key head in linear attention.

linear_value_head_dim (`int`, *optional*, defaults to 128) : Dimension of each value head in linear attention.

linear_num_key_heads (`int`, *optional*, defaults to 16) : Number of key heads used in linear attention layers.

linear_num_value_heads (`int`, *optional*, defaults to 32) : Number of value heads used in linear attention layers.

moe_intermediate_size (`int`, *optional*, defaults to `512`) : Intermediate size of the routed expert MLPs.

shared_expert_intermediate_size (`int`, *optional*, defaults to `512`) : Intermediate size of the shared expert MLPs.

num_experts_per_tok (`int`, *optional*, defaults to `10`) : Number of experts to route each token to. This is the top-k value for the token-choice routing.

num_experts (`int`, *optional*, defaults to `512`) : Number of routed experts in MoE layers. 

output_router_logits (`bool`, *optional*, defaults to `False`) : Whether or not the router logits should be returned by the model. Enabling this will also allow the model to output the auxiliary loss, including load balancing loss and router z-loss.

router_aux_loss_coef (`float`, *optional*, defaults to `0.001`) : Auxiliary load balancing loss coefficient. Used to penalize uneven expert routing in MoE models.

layer_types (`list[str]`, *optional*) : A list that explicitly maps each layer index with its layer type. If not provided, it will be automatically generated based on config values.

pad_token_id (`int`, *optional*) : Token id used for padding in the vocabulary.

bos_token_id (`int`, *optional*) : Token id used for beginning-of-stream in the vocabulary.

eos_token_id (`Union[int, list[int]]`, *optional*) : Token id used for end-of-stream in the vocabulary.

hc_count (`int`, *optional*, defaults to 4) : Number of residual streams used by the hyper-connections.

hc_lowrank (`int`, *optional*, defaults to 320) : Rank of the learned hyper-connection input mixer.

ple_layer_ids (`list[int]`, *optional*) : One-indexed decoder layer ids that use Per-Layer Embedding (PLE).

ple_embed_dim (`int`, *optional*) : Total dimension of the embeddings concatenated from all n-gram heads in each PLE module. Defaults to `hidden_size`.

ple_conv_kernel_size (`int`, *optional*, defaults to 4) : Kernel size of the dilated depthwise convolution in each PLE module.

ngram_size (`int`, *optional*, defaults to 3) : Largest token n-gram represented by PLE.

heads_per_ngram (`int`, *optional*, defaults to 8) : Number of independently hashed embedding heads for every n-gram order.

ngram_vocab_size_base (`int`, *optional*, defaults to 20000000) : Lower bound used to derive a distinct prime vocabulary size for each hashed n-gram head.

make_ngram_vocab_size_divisible_by (`int`, *optional*, defaults to 128) : Divisor used to pad the combined n-gram embedding vocabulary.

seed (`int`, *optional*, defaults to 1234) : Seed used to deterministically derive the per-layer n-gram hash multipliers.

split_ngram_parts (`int`, *optional*, defaults to 512) : Number of checkpoint shards used for each PLE n-gram embedding table. Loading concatenates the shards into a single runtime embedding, while `save_pretrained` restores the configured sharded layout.

indexer_n_heads (`int`, *optional*) : Number of query heads used by the QSA token indexer. Setting this enables QSA on full-attention layers.

indexer_kv_heads (`int`, *optional*) : Number of indexer key heads. Qwen4-Exp QSA requires one key head.

indexer_head_dim (`int`, *optional*) : Dimension of every QSA indexer query and key head.

indexer_budget (`int`, *optional*) : Maximum number of tokens selected from complete compressed blocks for each query.

indexer_compress_ratio (`int`, *optional*) : Number of consecutive token keys averaged into one QSA index block.

norm_topk_prob (`bool`, *optional*, defaults to `True`) : Whether to normalize the weights of the routed experts. 

output_gate_type (`str`, *optional*) : Activation used by the output gate of linear attention. If unset, `hidden_act` is used.

This is the configuration class to store the configuration of a Qwen4ExpModel. It is used to instantiate a Qwen4 Exp
model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
defaults will yield a similar configuration to that of the [Qwen/Qwen4-Exp](https://huggingface.co/Qwen/Qwen4-Exp)

Configuration objects inherit from [PreTrainedConfig](/docs/transformers/main/en/main_classes/configuration#transformers.PreTrainedConfig) and can be used to control the model outputs. Read the
documentation from [PreTrainedConfig](/docs/transformers/main/en/main_classes/configuration#transformers.PreTrainedConfig) for more information.

#### validate_architecture[[transformers.Qwen4ExpTextConfig.validate_architecture]]

```python
validate_architecture()
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/configuration_qwen4_exp.py#L188)

Part of `@strict`-powered validation. Validates Qwen4-Exp architecture invariants.

## Qwen4ExpVisionConfig[[transformers.Qwen4ExpVisionConfig]]

#### transformers.Qwen4ExpVisionConfig[[transformers.Qwen4ExpVisionConfig]]

```python
transformers.Qwen4ExpVisionConfig(transformers_version: str | None = None, architectures: list[str] | None = None, output_hidden_states: bool | None = False, return_dict: bool | None = True, dtype: typing.Union[str, ForwardRef('torch.dtype'), NoneType] = None, chunk_size_feed_forward: int = 0, is_encoder_decoder: bool = False, id2label: dict[int, str] | dict[str, str] | None = None, label2id: dict[str, int] | dict[str, str] | None = None, problem_type: typing.Optional[typing.Literal['regression', 'single_label_classification', 'multi_label_classification']] = None, depth: int = 27, hidden_size: int = 1152, hidden_act: str = 'gelu_pytorch_tanh', intermediate_size: int = 4304, num_heads: int = 16, in_channels: int = 3, patch_size: int | list[int] | tuple[int, int] = 16, spatial_merge_size: int = 2, temporal_patch_size: int | list[int] | tuple[int, int] = 2, out_hidden_size: int = 3584, num_position_embeddings: int = 2304, initializer_range: float = 0.02)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/configuration_qwen4_exp.py#L262)

**Parameters:**

depth (`int`, *optional*, defaults to `27`) : Number of Transformer layers in the vision encoder.

hidden_size (`int`, *optional*, defaults to `1152`) : Dimension of the hidden representations.

hidden_act (`str`, *optional*, defaults to `gelu_pytorch_tanh`) : The non-linear activation function (function or string) in the decoder. For example, `"gelu"`, `"relu"`, `"silu"`, etc.

intermediate_size (`int`, *optional*, defaults to `4304`) : Dimension of the MLP representations.

num_heads (`int`, *optional*, defaults to `16`) : Number of attention heads for each attention layer in the Transformer decoder.

in_channels (`int`, *optional*, defaults to `3`) : The number of input channels.

patch_size (`Union[int, list[int], tuple[int, int]]`, *optional*, defaults to `16`) : The size (resolution) of each patch.

spatial_merge_size (`int`, *optional*, defaults to `2`) : The size of the spatial merge window used to reduce the number of visual tokens by merging neighboring patches.

temporal_patch_size (`Union[int, list[int], tuple[int, int]]`, *optional*, defaults to `2`) : Temporal patch size used in the 3D patch embedding for video inputs.

out_hidden_size (`int`, *optional*, defaults to 3584) : The output hidden size of the vision model.

num_position_embeddings (`int`, *optional*, defaults to 2304) : The maximum sequence length that this model might ever be used with

initializer_range (`float`, *optional*, defaults to `0.02`) : The standard deviation of the truncated_normal_initializer for initializing all weight matrices.

This is the configuration class to store the configuration of a Qwen4ExpModel. It is used to instantiate a Qwen4 Exp
model according to the specified arguments, defining the model architecture. Instantiating a configuration with the
defaults will yield a similar configuration to that of the [Qwen/Qwen4-Exp](https://huggingface.co/Qwen/Qwen4-Exp)

Configuration objects inherit from [PreTrainedConfig](/docs/transformers/main/en/main_classes/configuration#transformers.PreTrainedConfig) and can be used to control the model outputs. Read the
documentation from [PreTrainedConfig](/docs/transformers/main/en/main_classes/configuration#transformers.PreTrainedConfig) for more information.

## Qwen4ExpVisionModel[[transformers.Qwen4ExpVisionModel]]

#### transformers.Qwen4ExpVisionModel[[transformers.Qwen4ExpVisionModel]]

```python
transformers.Qwen4ExpVisionModel(config, *inputs, **kwargs)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L1849)

**Parameters:**

config ([Qwen4ExpVisionModel](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpVisionModel)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights.

The bare Qwen4 Exp Model outputting raw hidden-states without any specific head on top.

This model inherits from [PreTrainedModel](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the
library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
etc.)

This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
and behavior.

#### forward[[transformers.Qwen4ExpVisionModel.forward]]

```python
forward(hidden_states: Tensor, grid_thw: Tensor, **kwargs)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L1912)

**Parameters:**

hidden_states (`torch.Tensor` of shape `(seq_len, hidden_size)`) : The final hidden states of the model.

grid_thw (`torch.Tensor` of shape `(num_images_or_videos, 3)`) : The temporal, height and width of feature shape of each image in LLM.

**Returns:** `torch.Tensor`

hidden_states.

## Qwen4ExpTextModel[[transformers.Qwen4ExpTextModel]]

#### transformers.Qwen4ExpTextModel[[transformers.Qwen4ExpTextModel]]

```python
transformers.Qwen4ExpTextModel(config: Qwen4ExpTextConfig)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L1314)

**Parameters:**

config ([Qwen4ExpTextConfig](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpTextConfig)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights.

The bare Qwen4 Exp Text Model outputting raw hidden-states without any specific head on top.

This model inherits from [PreTrainedModel](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the
library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
etc.)

This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
and behavior.

#### forward[[transformers.Qwen4ExpTextModel.forward]]

```python
forward(input_ids: typing.Optional[torch.LongTensor] = None, attention_mask: typing.Optional[torch.Tensor] = None, position_ids: typing.Optional[torch.LongTensor] = None, past_key_values: transformers.cache_utils.Cache | None = None, inputs_embeds: typing.Optional[torch.FloatTensor] = None, use_cache: bool | None = None, ple_input_ids: typing.Optional[torch.Tensor] = None, **kwargs: Unpack)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L1334)

**Parameters:**

input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of input sequence tokens in the vocabulary. Padding will be ignored by default.  Indices can be obtained using [AutoTokenizer](/docs/transformers/main/en/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/main/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/main/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details.  [What are input IDs?](../glossary#input-ids)

attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) : Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:  - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**.  [What are attention masks?](../glossary#attention-mask)

position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`.  [What are position IDs?](../glossary#position-ids)

past_key_values (`~cache_utils.Cache`, *optional*) : Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values` returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`.  Only [Cache](/docs/transformers/main/en/internal/generation_utils#transformers.Cache) instance is allowed as input, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache). If no `past_key_values` are passed, [DynamicCache](/docs/transformers/main/en/internal/generation_utils#transformers.DynamicCache) will be initialized by default.  The model will output the same cache format that is fed as input.  If `past_key_values` are used, the user is expected to input only unprocessed `input_ids` (those that don't have their past key value states given to this model) of shape `(batch_size, unprocessed_length)` instead of all `input_ids` of shape `(batch_size, sequence_length)`.

inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) : Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix.

use_cache (`bool`, *optional*) : If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see `past_key_values`).

ple_input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Original token ids used by Per-Layer Embedding (PLE). This is only needed when PLE is enabled and `inputs_embeds` are passed instead of `input_ids`.

**Returns:** [BaseModelOutputWithPast](/docs/transformers/main/en/main_classes/output#transformers.modeling_outputs.BaseModelOutputWithPast) or `tuple(torch.FloatTensor)`

A [BaseModelOutputWithPast](/docs/transformers/main/en/main_classes/output#transformers.modeling_outputs.BaseModelOutputWithPast) or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration ([Qwen4ExpConfig](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpConfig)) and inputs.

The [Qwen4ExpTextModel](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpTextModel) forward method, overrides the `__call__` special method.

Although the recipe for forward pass needs to be defined within this function, one should call the `Module`
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.

- **last_hidden_state** (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`) -- Sequence of hidden-states at the output of the last layer of the model.

  If `past_key_values` is used only the last hidden-state of the sequences of shape `(batch_size, 1,
  hidden_size)` is output.
- **past_key_values** (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`) -- It is a [Cache](/docs/transformers/main/en/internal/generation_utils#transformers.Cache) instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

  Contains pre-computed hidden-states (key and values in the self-attention blocks and optionally if
  `config.is_encoder_decoder=True` in the cross-attention blocks) that can be used (see `past_key_values`
  input) to speed up sequential decoding.
- **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
  one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

  Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
- **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
  sequence_length)`.

  Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
  heads.

## Qwen4ExpModel[[transformers.Qwen4ExpModel]]

#### transformers.Qwen4ExpModel[[transformers.Qwen4ExpModel]]

```python
transformers.Qwen4ExpModel(config)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L1965)

**Parameters:**

config ([Qwen4ExpModel](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpModel)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights.

The bare Qwen4 Exp Model outputting raw hidden-states without any specific head on top.

This model inherits from [PreTrainedModel](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the
library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
etc.)

This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
and behavior.

#### forward[[transformers.Qwen4ExpModel.forward]]

```python
forward(input_ids: LongTensor = None, attention_mask: typing.Optional[torch.Tensor] = None, position_ids: typing.Optional[torch.LongTensor] = None, past_key_values: transformers.cache_utils.Cache | None = None, inputs_embeds: typing.Optional[torch.FloatTensor] = None, pixel_values: typing.Optional[torch.Tensor] = None, pixel_values_videos: typing.Optional[torch.FloatTensor] = None, image_grid_thw: typing.Optional[torch.LongTensor] = None, video_grid_thw: typing.Optional[torch.LongTensor] = None, mm_token_type_ids: typing.Optional[torch.IntTensor] = None, **kwargs: Unpack)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L2247)

**Parameters:**

input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of input sequence tokens in the vocabulary. Padding will be ignored by default.  Indices can be obtained using [AutoTokenizer](/docs/transformers/main/en/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/main/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/main/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details.  [What are input IDs?](../glossary#input-ids)

attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) : Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:  - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**.  [What are attention masks?](../glossary#attention-mask)

position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`.  [What are position IDs?](../glossary#position-ids)

past_key_values (`~cache_utils.Cache`, *optional*) : Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values` returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`.  Only [Cache](/docs/transformers/main/en/internal/generation_utils#transformers.Cache) instance is allowed as input, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache). If no `past_key_values` are passed, [DynamicCache](/docs/transformers/main/en/internal/generation_utils#transformers.DynamicCache) will be initialized by default.  The model will output the same cache format that is fed as input.  If `past_key_values` are used, the user is expected to input only unprocessed `input_ids` (those that don't have their past key value states given to this model) of shape `(batch_size, unprocessed_length)` instead of all `input_ids` of shape `(batch_size, sequence_length)`.

inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) : Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix.

pixel_values (`torch.Tensor` of shape `(batch_size, num_channels, image_size, image_size)`, *optional*) : The tensors corresponding to the input images. Pixel values can be obtained using `image_processor_class`. See `image_processor_class.__call__` for details (`processor_class` uses `image_processor_class` for processing images).

pixel_values_videos (`torch.FloatTensor` of shape `(batch_size, num_frames, num_channels, frame_size, frame_size)`, *optional*) : The tensors corresponding to the input video. Pixel values for videos can be obtained using `video_processor_class`. See `video_processor_class.__call__` for details (`processor_class` uses `video_processor_class` for processing videos).

image_grid_thw (`torch.LongTensor` of shape `(num_images, 3)`, *optional*) : The temporal, height and width of feature shape of each image in LLM.

video_grid_thw (`torch.LongTensor` of shape `(num_videos, 3)`, *optional*) : The temporal, height and width of feature shape of each video in LLM.

mm_token_type_ids (`torch.IntTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of input sequence tokens matching each modality. For example text (0), image (1), video (2). Multimodal token type ids can be obtained using [AutoProcessor](/docs/transformers/main/en/model_doc/auto#transformers.AutoProcessor). See [ProcessorMixin.__call__()](/docs/transformers/main/en/main_classes/processors#transformers.ProcessorMixin.__call__) for details.

**Returns:** `Qwen4ExpModelOutputWithPast` or `tuple(torch.FloatTensor)`

A `Qwen4ExpModelOutputWithPast` or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration (`None`) and inputs.

The [Qwen4ExpModel](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpModel) forward method, overrides the `__call__` special method.

Although the recipe for forward pass needs to be defined within this function, one should call the `Module`
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.

- **last_hidden_state** (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`) -- Sequence of hidden-states at the output of the last layer of the model.

  If `past_key_values` is used only the last hidden-state of the sequences of shape `(batch_size, 1,
  hidden_size)` is output.
- **past_key_values** (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`) -- It is a [Cache](/docs/transformers/main/en/internal/generation_utils#transformers.Cache) instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

  Contains pre-computed hidden-states (key and values in the self-attention blocks and optionally if
  `config.is_encoder_decoder=True` in the cross-attention blocks) that can be used (see `past_key_values`
  input) to speed up sequential decoding.
- **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
  one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

  Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
- **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
  sequence_length)`.

  Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
  heads.
- **rope_deltas** (`torch.LongTensor` of shape `(batch_size, )`, *optional*) -- The rope index difference between sequence length and multimodal rope.
  The attribute is deprecated and will be removed in v5.20, use `model.base_model.rope_deltas` instead.
- **router_logits** (`tuple[torch.FloatTensor]`, *optional*, returned when `output_router_logits=True` is passed or when `config.add_router_probs=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, sequence_length, num_experts)`.

  Router logits of the model, useful to compute the auxiliary loss for Mixture of Experts models.

## Qwen4ExpForCausalLM[[transformers.Qwen4ExpForCausalLM]]

#### transformers.Qwen4ExpForCausalLM[[transformers.Qwen4ExpForCausalLM]]

```python
transformers.Qwen4ExpForCausalLM(config)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L1529)

**Parameters:**

config ([Qwen4ExpForCausalLM](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpForCausalLM)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights.

The Qwen4 Exp Model for causal language modeling.

This model inherits from [PreTrainedModel](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the
library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
etc.)

This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
and behavior.

#### forward[[transformers.Qwen4ExpForCausalLM.forward]]

```python
forward(input_ids: typing.Optional[torch.LongTensor] = None, attention_mask: typing.Optional[torch.Tensor] = None, position_ids: typing.Optional[torch.LongTensor] = None, past_key_values: transformers.cache_utils.Cache | None = None, inputs_embeds: typing.Optional[torch.FloatTensor] = None, labels: typing.Optional[torch.LongTensor] = None, use_cache: bool | None = None, output_router_logits: bool | None = None, logits_to_keep: typing.Union[int, torch.Tensor] = 0, **kwargs: Unpack)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L1549)

**Parameters:**

input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of input sequence tokens in the vocabulary. Padding will be ignored by default.  Indices can be obtained using [AutoTokenizer](/docs/transformers/main/en/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/main/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/main/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details.  [What are input IDs?](../glossary#input-ids)

attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) : Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:  - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**.  [What are attention masks?](../glossary#attention-mask)

position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`.  [What are position IDs?](../glossary#position-ids)

past_key_values (`~cache_utils.Cache`, *optional*) : Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values` returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`.  Only [Cache](/docs/transformers/main/en/internal/generation_utils#transformers.Cache) instance is allowed as input, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache). If no `past_key_values` are passed, [DynamicCache](/docs/transformers/main/en/internal/generation_utils#transformers.DynamicCache) will be initialized by default.  The model will output the same cache format that is fed as input.  If `past_key_values` are used, the user is expected to input only unprocessed `input_ids` (those that don't have their past key value states given to this model) of shape `(batch_size, unprocessed_length)` instead of all `input_ids` of shape `(batch_size, sequence_length)`.

inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) : Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix.

labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Labels for computing the masked language modeling loss. Indices should either be in `[0, ..., config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.

use_cache (`bool`, *optional*) : If set to `True`, `past_key_values` key value states are returned and can be used to speed up decoding (see `past_key_values`).

output_router_logits (`bool`, *optional*) : Whether or not to return the logits of all the routers. They are useful for computing the router loss, and should not be returned during inference.

logits_to_keep (`Union[int, torch.Tensor]`, *optional*, defaults to `0`) : If an `int`, compute logits for the last `logits_to_keep` tokens. If `0`, calculate logits for all `input_ids` (special case). Only last token logits are needed for generation, and calculating them only for that token can save memory, which becomes pretty significant for long sequences or large vocabulary size. If a `torch.Tensor`, must be 1D corresponding to the indices to keep in the sequence length dimension. This is useful when using packed tensor format (single dimension for batch and sequence length).

**Returns:** `MoeCausalLMOutputWithPast` or `tuple(torch.FloatTensor)`

A `MoeCausalLMOutputWithPast` or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration ([Qwen4ExpConfig](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpConfig)) and inputs.

The [Qwen4ExpForCausalLM](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpForCausalLM) forward method, overrides the `__call__` special method.

Although the recipe for forward pass needs to be defined within this function, one should call the `Module`
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.

- **loss** (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided) -- Language modeling loss (for next-token prediction).
- **logits** (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`) -- Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
- **aux_loss** (`torch.FloatTensor`, *optional*, returned when `labels` is provided) -- aux_loss for the sparse modules.
- **router_logits** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_router_probs=True` and `config.add_router_probs=True` is passed or when `config.output_router_probs=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, sequence_length, num_experts)`.

  Raw router logits (post-softmax) that are computed by MoE routers, these terms are used to compute the auxiliary
  loss for Mixture of Experts models.
- **past_key_values** (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`) -- It is a [Cache](/docs/transformers/main/en/internal/generation_utils#transformers.Cache) instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

  Contains pre-computed hidden-states (key and values in the self-attention blocks) that can be used (see
  `past_key_values` input) to speed up sequential decoding.
- **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
  one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

  Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
- **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
  sequence_length)`.

  Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
  heads.

Example:

```python
>>> from transformers import AutoTokenizer, Qwen4ExpForCausalLM

>>> model = Qwen4ExpForCausalLM.from_pretrained("Qwen/Qwen3-Next-80B-A3B-Instruct")
>>> tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen3-Next-80B-A3B-Instruct")

>>> prompt = "Hey, are you conscious? Can you talk to me?"
>>> inputs = tokenizer(prompt, return_tensors="pt")

>>> # Generate
>>> generate_ids = model.generate(inputs.input_ids, max_length=30)
>>> tokenizer.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
"Hey, are you conscious? Can you talk to me?\nI'm not conscious, but I can talk to you."
```

## Qwen4ExpForConditionalGeneration[[transformers.Qwen4ExpForConditionalGeneration]]

#### transformers.Qwen4ExpForConditionalGeneration[[transformers.Qwen4ExpForConditionalGeneration]]

```python
transformers.Qwen4ExpForConditionalGeneration(config)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L2341)

**Parameters:**

config ([Qwen4ExpForConditionalGeneration](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpForConditionalGeneration)) : Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the [from_pretrained()](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights.

The Qwen4 Exp Model for token generation conditioned on other modalities (e.g. image-text-to-text generation).

This model inherits from [PreTrainedModel](/docs/transformers/main/en/main_classes/model#transformers.PreTrainedModel). Check the superclass documentation for the generic methods the
library implements for all its model (such as downloading or saving, resizing the input embeddings, pruning heads
etc.)

This model is also a PyTorch [torch.nn.Module](https://pytorch.org/docs/stable/nn.html#torch.nn.Module) subclass.
Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage
and behavior.

#### forward[[transformers.Qwen4ExpForConditionalGeneration.forward]]

```python
forward(input_ids: LongTensor = None, attention_mask: typing.Optional[torch.Tensor] = None, position_ids: typing.Optional[torch.LongTensor] = None, past_key_values: transformers.cache_utils.Cache | None = None, inputs_embeds: typing.Optional[torch.FloatTensor] = None, labels: typing.Optional[torch.LongTensor] = None, pixel_values: typing.Optional[torch.Tensor] = None, pixel_values_videos: typing.Optional[torch.FloatTensor] = None, image_grid_thw: typing.Optional[torch.LongTensor] = None, video_grid_thw: typing.Optional[torch.LongTensor] = None, mm_token_type_ids: typing.Optional[torch.IntTensor] = None, logits_to_keep: typing.Union[int, torch.Tensor] = 0, **kwargs: Unpack)
```

[Source](https://github.com/huggingface/transformers/blob/main/src/transformers/models/qwen4_exp/modeling_qwen4_exp.py#L2382)

**Parameters:**

input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of input sequence tokens in the vocabulary. Padding will be ignored by default.  Indices can be obtained using [AutoTokenizer](/docs/transformers/main/en/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/main/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/main/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details.  [What are input IDs?](../glossary#input-ids)

attention_mask (`torch.Tensor` of shape `(batch_size, sequence_length)`, *optional*) : Mask to avoid performing attention on padding token indices. Mask values selected in `[0, 1]`:  - 1 for tokens that are **not masked**, - 0 for tokens that are **masked**.  [What are attention masks?](../glossary#attention-mask)

position_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of positions of each input sequence tokens in the position embeddings. Selected in the range `[0, config.n_positions - 1]`.  [What are position IDs?](../glossary#position-ids)

past_key_values (`~cache_utils.Cache`, *optional*) : Pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used to speed up sequential decoding. This typically consists in the `past_key_values` returned by the model at a previous stage of decoding, when `use_cache=True` or `config.use_cache=True`.  Only [Cache](/docs/transformers/main/en/internal/generation_utils#transformers.Cache) instance is allowed as input, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache). If no `past_key_values` are passed, [DynamicCache](/docs/transformers/main/en/internal/generation_utils#transformers.DynamicCache) will be initialized by default.  The model will output the same cache format that is fed as input.  If `past_key_values` are used, the user is expected to input only unprocessed `input_ids` (those that don't have their past key value states given to this model) of shape `(batch_size, unprocessed_length)` instead of all `input_ids` of shape `(batch_size, sequence_length)`.

inputs_embeds (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*) : Optionally, instead of passing `input_ids` you can choose to directly pass an embedded representation. This is useful if you want more control over how to convert `input_ids` indices into associated vectors than the model's internal embedding lookup matrix.

labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*) : Labels for computing the masked language modeling loss. Indices should either be in `[0, ..., config.vocab_size]` or -100 (see `input_ids` docstring). Tokens with indices set to `-100` are ignored (masked), the loss is only computed for the tokens with labels in `[0, ..., config.vocab_size]`.

pixel_values (`torch.Tensor` of shape `(batch_size, num_channels, image_size, image_size)`, *optional*) : The tensors corresponding to the input images. Pixel values can be obtained using [Qwen2VLImageProcessor](/docs/transformers/main/en/model_doc/qwen2_vl#transformers.Qwen2VLImageProcessor). See `Qwen2VLImageProcessor.__call__()` for details ([Qwen3VLProcessor](/docs/transformers/main/en/model_doc/qwen3_vl#transformers.Qwen3VLProcessor) uses [Qwen2VLImageProcessor](/docs/transformers/main/en/model_doc/qwen2_vl#transformers.Qwen2VLImageProcessor) for processing images).

pixel_values_videos (`torch.FloatTensor` of shape `(batch_size, num_frames, num_channels, frame_size, frame_size)`, *optional*) : The tensors corresponding to the input video. Pixel values for videos can be obtained using [Qwen3VLVideoProcessor](/docs/transformers/main/en/model_doc/qwen3_vl#transformers.Qwen3VLVideoProcessor). See `Qwen3VLVideoProcessor.__call__()` for details ([Qwen3VLProcessor](/docs/transformers/main/en/model_doc/qwen3_vl#transformers.Qwen3VLProcessor) uses [Qwen3VLVideoProcessor](/docs/transformers/main/en/model_doc/qwen3_vl#transformers.Qwen3VLVideoProcessor) for processing videos).

image_grid_thw (`torch.LongTensor` of shape `(num_images, 3)`, *optional*) : The temporal, height and width of feature shape of each image in LLM.

video_grid_thw (`torch.LongTensor` of shape `(num_videos, 3)`, *optional*) : The temporal, height and width of feature shape of each video in LLM.

mm_token_type_ids (`torch.IntTensor` of shape `(batch_size, sequence_length)`, *optional*) : Indices of input sequence tokens matching each modality. For example text (0), image (1), video (2). Multimodal token type ids can be obtained using [AutoProcessor](/docs/transformers/main/en/model_doc/auto#transformers.AutoProcessor). See [ProcessorMixin.__call__()](/docs/transformers/main/en/main_classes/processors#transformers.ProcessorMixin.__call__) for details. 

logits_to_keep (`Union[int, torch.Tensor]`, *optional*, defaults to `0`) : If an `int`, compute logits for the last `logits_to_keep` tokens. If `0`, calculate logits for all `input_ids` (special case). Only last token logits are needed for generation, and calculating them only for that token can save memory, which becomes pretty significant for long sequences or large vocabulary size. If a `torch.Tensor`, must be 1D corresponding to the indices to keep in the sequence length dimension. This is useful when using packed tensor format (single dimension for batch and sequence length).

**Returns:** `Qwen4ExpCausalLMOutputWithPast` or `tuple(torch.FloatTensor)`

A `Qwen4ExpCausalLMOutputWithPast` or a tuple of
`torch.FloatTensor` (if `return_dict=False` is passed or when `config.return_dict=False`) comprising various
elements depending on the configuration ([Qwen4ExpConfig](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpConfig)) and inputs.

The [Qwen4ExpForConditionalGeneration](/docs/transformers/main/en/model_doc/qwen4_exp#transformers.Qwen4ExpForConditionalGeneration) forward method, overrides the `__call__` special method.

Although the recipe for forward pass needs to be defined within this function, one should call the `Module`
instance afterwards instead of this since the former takes care of running the pre and post processing steps while
the latter silently ignores them.

- **loss** (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` is provided) -- Language modeling loss (for next-token prediction).
- **logits** (`torch.FloatTensor` of shape `(batch_size, sequence_length, config.vocab_size)`) -- Prediction scores of the language modeling head (scores for each vocabulary token before SoftMax).
- **past_key_values** (`Cache`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`) -- It is a [Cache](/docs/transformers/main/en/internal/generation_utils#transformers.Cache) instance. For more details, see our [kv cache guide](https://huggingface.co/docs/transformers/en/kv_cache).

  Contains pre-computed hidden-states (key and values in the self-attention blocks) that can be used (see
  `past_key_values` input) to speed up sequential decoding.
- **hidden_states** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_hidden_states=True` is passed or when `config.output_hidden_states=True`) -- Tuple of `torch.FloatTensor` (one for the output of the embeddings, if the model has an embedding layer, +
  one for the output of each layer) of shape `(batch_size, sequence_length, hidden_size)`.

  Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.
- **attentions** (`tuple(torch.FloatTensor)`, *optional*, returned when `output_attentions=True` is passed or when `config.output_attentions=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, num_heads, sequence_length,
  sequence_length)`.

  Attentions weights after the attention softmax, used to compute the weighted average in the self-attention
  heads.
- **rope_deltas** (`torch.LongTensor` of shape `(batch_size, )`, *optional*) -- The rope index difference between sequence length and multimodal rope.
  The attribute is deprecated and will be removed in v5.20, use `model.base_model.rope_deltas` instead.
- **router_logits** (`tuple[torch.FloatTensor]`, *optional*, returned when `output_router_logits=True` is passed or when `config.add_router_probs=True`) -- Tuple of `torch.FloatTensor` (one for each layer) of shape `(batch_size, sequence_length, num_experts)`.

  Router logits of the model, useful to compute the auxiliary loss for Mixture of Experts models.
- **aux_loss** (`torch.FloatTensor`, *optional*, returned when `labels` is provided) -- aux_loss for the sparse modules.

Example:
```python
>>> from transformers import AutoProcessor, Qwen4ExpForConditionalGeneration

>>> model = Qwen4ExpForConditionalGeneration.from_pretrained("Qwen/Qwen3.5-35B-A3B-Instruct", dtype="auto", device_map="auto")
>>> processor = AutoProcessor.from_pretrained("Qwen/Qwen3.5-35B-A3B-Instruct")

>>> messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "image",
                "image": "https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-VL/assets/demo.jpeg",
            },
            {"type": "text", "text": "Describe this image in short."},
        ],
    }
]

>>> # Preparation for inference
>>> inputs = processor.apply_chat_template(
    messages,
    tokenize=True,
    add_generation_prompt=True,
    return_dict=True,
    return_tensors="pt"
)
>>> inputs = inputs.to(model.device)

>>> # Generate
>>> generated_ids = model.generate(**inputs, max_new_tokens=128)
>>> generated_ids_trimmed = [
    out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
]
>>> processor.batch_decode(generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0]
"A woman in a plaid shirt sits on a sandy beach at sunset, smiling as she gives a high-five to a yellow Labrador Retriever wearing a harness. The ocean waves roll in the background."
```