# Emu3 ## Overview The Emu3 model was proposed in [Emu3: Next-Token Prediction is All You Need](https://huggingface.co/papers/2409.18869) by Xinlong Wang, Xiaosong Zhang, Zhengxiong Luo, Quan Sun, Yufeng Cui, Jinsheng Wang, Fan Zhang, Yueze Wang, Zhen Li, Qiying Yu, Yingli Zhao, Yulong Ao, Xuebin Min, Tao Li, Boya Wu, Bo Zhao, Bowen Zhang, Liangdong Wang, Guang Liu, Zheqi He, Xi Yang, Jingjing Liu, Yonghua Lin, Tiejun Huang, Zhongyuan Wang. Emu3 is a multimodal LLM that uses vector quantization to tokenize images into discrete tokens. Discretized image tokens are later fused with text token ids for image and text generation. The model can additionally generate images by predicting image token ids. The abstract from the paper is the following: *While next-token prediction is considered a promising path towards artificial general intelligence, it has struggled to excel in multimodal tasks, which are still dominated by diffusion models (e.g., Stable Diffusion) and compositional approaches (e.g., CLIP combined with LLMs). In this paper, we introduce Emu3, a new suite of state-of-the-art multimodal models trained solely with next-token prediction. By tokenizing images, text, and videos into a discrete space, we train a single transformer from scratch on a mixture of multimodal sequences. Emu3 outperforms several well-established task-specific models in both generation and perception tasks, surpassing flagship models such as SDXL and LLaVA-1.6, while eliminating the need for diffusion or compositional architectures. Emu3 is also capable of generating high-fidelity video via predicting the next token in a video sequence. We simplify complex multimodal model designs by converging on a singular focus: tokens, unlocking great potential for scaling both during training and inference. Our results demonstrate that next-token prediction is a promising path towards building general multimodal intelligence beyond language. We open-source key techniques and models to support further research in this direction.* Tips: - We advise users to set `processor.tokenizer.padding_side = "left"` before batched generation as it leads to more accurate results. - Note that the model has been trained with a specific prompt format for chatting. Use `processor.apply_chat_template(my_conversation_dict)` to correctly format your prompts. - Emu3 has two different checkpoints for image-generation and text-generation, make sure to use the correct checkpoint when loading the model. To generate an image, it is advised to use `prefix_constraints` so that the generated tokens are sampled only from possible image tokens. See more below for usage examples. > [!TIP] > Emu3 implementation in Transformers uses a special image token to indicate where to merge image embeddings. The special image token isn't new and uses one of the reserved tokens: ``. You have to add `` to your prompt in the place where the image should be embedded for correct generation. This model was contributed by [RaushanTurganbay](https://huggingface.co/RaushanTurganbay). The original code can be found [here](https://github.com/baaivision/Emu3). ## Usage example ### Text generation inference Here's how to load the model and perform inference in half-precision (`torch.bfloat16`) to generate textual output from text or text and image inputs: ```python from transformers import Emu3Processor, Emu3ForConditionalGeneration import torch from PIL import Image import requests processor = Emu3Processor.from_pretrained("BAAI/Emu3-Chat-hf") model = Emu3ForConditionalGeneration.from_pretrained("BAAI/Emu3-Chat-hf", dtype=torch.bfloat16, device_map="auto") # prepare image and text prompt url = 'http://images.cocodataset.org/val2017/000000039769.jpg' image = Image.open(requests.get(url, stream=True).raw) prompt = "What do you see in this image?" inputs = processor(images=image, text=prompt, return_tensors="pt").to(model.device, dtype=torch.bfloat16) # autoregressively complete prompt output = model.generate(**inputs, max_new_tokens=50) print(processor.decode(output[0], skip_special_tokens=True)) ``` ### Image generation inference Emu3 can also generate images from textual input. Here is how you can do it: ```python processor = Emu3Processor.from_pretrained("BAAI/Emu3-Gen-hf") model = Emu3ForConditionalGeneration.from_pretrained("BAAI/Emu3-Gen-hf", dtype="bfloat16", device_map="auto", attn_implementation="flash_attention_2") inputs = processor( text=["a portrait of young girl. masterpiece, film grained, best quality.", "a dog running under the rain"], padding=True, return_tensors="pt", return_for_image_generation=True, ) inputs = inputs.to(device=model.device, dtype=torch.bfloat16) neg_prompt = "lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, normal quality, jpeg artifacts, signature, watermark, username, blurry." neg_inputs = processor(text=[neg_prompt] * 2, return_tensors="pt").to(device=model.device) image_sizes = inputs.pop("image_sizes") HEIGHT, WIDTH = image_sizes[0] VISUAL_TOKENS = model.vocabulary_mapping.image_tokens def prefix_allowed_tokens_fn(batch_id, input_ids): height, width = HEIGHT, WIDTH visual_tokens = VISUAL_TOKENS image_wrapper_token_id = torch.tensor([processor.tokenizer.image_wrapper_token_id], device=model.device) eoi_token_id = torch.tensor([processor.tokenizer.eoi_token_id], device=model.device) eos_token_id = torch.tensor([processor.tokenizer.eos_token_id], device=model.device) pad_token_id = torch.tensor([processor.tokenizer.pad_token_id], device=model.device) eof_token_id = torch.tensor([processor.tokenizer.eof_token_id], device=model.device) eol_token_id = processor.tokenizer.encode("", return_tensors="pt")[0] position = torch.nonzero(input_ids == image_wrapper_token_id, as_tuple=True)[0][0] offset = input_ids.shape[0] - position if offset % (width + 1) == 0: return (eol_token_id, ) elif offset == (width + 1) * height + 1: return (eof_token_id, ) elif offset == (width + 1) * height + 2: return (eoi_token_id, ) elif offset == (width + 1) * height + 3: return (eos_token_id, ) elif offset > (width + 1) * height + 3: return (pad_token_id, ) else: return visual_tokens out = model.generate( **inputs, max_new_tokens=50_000, # make sure to have enough tokens for one image prefix_allowed_tokens_fn=prefix_allowed_tokens_fn, return_dict_in_generate=True, negative_prompt_ids=neg_inputs.input_ids, # indicate for Classifier-Free Guidance negative_prompt_attention_mask=neg_inputs.attention_mask, ) image = model.decode_image_tokens(out.sequences[:, inputs.input_ids.shape[1]: ], height=HEIGHT, width=WIDTH) images = processor.postprocess(list(image.float()), return_tensors="PIL.Image.Image") # internally we convert to np but it's not supported in bf16 precision for i, image in enumerate(images['pixel_values']): image.save(f"result{i}.png") ``` ## Emu3Config[[transformers.Emu3Config]] #### transformers.Emu3Config[[transformers.Emu3Config]] [Source](https://github.com/huggingface/transformers/blob/v4.57.1/src/transformers/models/emu3/configuration_emu3.py#L279) This is the configuration class to store the configuration of a [Emu3Model](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3Model). It is used to instantiate a emu3 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 [Emu3-community/Emu3-Chat-hf](https://huggingface.co/Emu3-community/Emu3-Chat-hf). Configuration objects inherit from [PretrainedConfig](/docs/transformers/v4.57.1/en/main_classes/configuration#transformers.PretrainedConfig) and can be used to control the model outputs. Read the documentation from [PretrainedConfig](/docs/transformers/v4.57.1/en/main_classes/configuration#transformers.PretrainedConfig) for more information. **Parameters:** vq_config (`Union[Dict, Emu3VQVAEConfig]`, *optional*) : Emu3VQVAEConfig instance containing the configuration for the VQ-VAE model. text_config (`Union[Dict, Emu3TextConfig]``, *optional*) : Emu3TextConfig instance containing the configuration for the language model. vocabulary_map (`dict`, *optional*) : A dictionary containing the vocabulary map from the tokenizer. Used to obtain tokens from the image inputs. ## Emu3VQVAEConfig[[transformers.Emu3VQVAEConfig]] #### transformers.Emu3VQVAEConfig[[transformers.Emu3VQVAEConfig]] [Source](https://github.com/huggingface/transformers/blob/v4.57.1/src/transformers/models/emu3/configuration_emu3.py#L23) This is the configuration class to store the configuration of a [Emu3VQVAE](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3VQVAE). It is used to instantiate an VQ-VAE model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a configuration to the VQ model presented in Emu3 paper. Configuration objects inherit from [PretrainedConfig](/docs/transformers/v4.57.1/en/main_classes/configuration#transformers.PretrainedConfig) and can be used to control the model outputs. Read the documentation from [PretrainedConfig](/docs/transformers/v4.57.1/en/main_classes/configuration#transformers.PretrainedConfig) for more information. ```python >>> from transformers import Emu3VQVAE, Emu3VQVAEConfig >>> # Initializing a video VQ model of Emu3 configuration >>> configuration = Emu3VQVAEConfig() >>> # Initializing a model from the Emu3 VQ model style configuration >>> model = Emu3VQVAE(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` **Parameters:** codebook_size (`int`, *optional*, defaults to 32768) : Codebook size of the VQ model. embed_dim (`int`, *optional*, defaults to 4) : Dimension of the quantized vector in codebook. latent_channels (`int`, *optional*, defaults to 4) : Dimension of the output channel of encoder and the input channel of decoder double_latent (`bool`, *optional*, defaults to `False`) : Whether double the output dim of the encoder. in_channels (`int`, *optional*, defaults to 3) : Input channel of encoder. out_channels (`int`, *optional*, defaults to 3) : Output channel of decoder. temporal_downsample_factor (`int`, *optional*, defaults to 4) : Temporal downsample factor. base_channels (`int`, *optional*, defaults to 256) : Basic channel number of the intermediate blocks. channel_multiplier (`list[int]`, *optional*, defaults to `[1, 2, 2, 4]`) : Channel scaling factor of the intermediate blocks. num_res_blocks (`int`, *optional*, defaults to 2) : Residual block number in each stage. attn_resolutions (`list[int]`, *optional*, defaults to `[3]`) : Stage indices to apply attention. hidden_size (`int`, *optional*, defaults to 1024) : Dimension of the hidden representations in the attention layer. num_attention_heads (`int`, *optional*, defaults to 1) : Number of attention heads for each attention layer. attention_dropout (`float`, *optional*, defaults to 0.0) : The dropout ratio for the attention probabilities. ## Emu3TextConfig[[transformers.Emu3TextConfig]] #### transformers.Emu3TextConfig[[transformers.Emu3TextConfig]] [Source](https://github.com/huggingface/transformers/blob/v4.57.1/src/transformers/models/emu3/configuration_emu3.py#L113) This is the configuration class to store the configuration of a [Emu3TextModel](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3TextModel). It is used to instantiate a emu3 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 [Emu3-community/Emu3-Chat-hf](https://huggingface.co/Emu3-community/Emu3-Chat-hf). Configuration objects inherit from [PretrainedConfig](/docs/transformers/v4.57.1/en/main_classes/configuration#transformers.PretrainedConfig) and can be used to control the model outputs. Read the documentation from [PretrainedConfig](/docs/transformers/v4.57.1/en/main_classes/configuration#transformers.PretrainedConfig) for more information. ```python >>> from transformers import Emu3Model, Emu3Config >>> # Initializing a Emu3-community/Emu3-Chat-hf style configuration >>> configuration = Emu3Config() >>> # Initializing a model from the Emu3-community/Emu3-Chat-hf style configuration >>> model = Emu3Model(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` **Parameters:** vocab_size (`int`, *optional*, defaults to 184622) : Vocabulary size of the Emu3 model. Defines the number of different tokens that can be represented by the `inputs_ids` passed when calling [Emu3Model](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3Model) hidden_size (`int`, *optional*, defaults to 4096) : Dimension of the hidden representations. intermediate_size (`int`, *optional*, defaults to 14336) : Dimension of the MLP representations. num_hidden_layers (`int`, *optional*, defaults to 32) : Number of hidden layers in the Transformer decoder. num_attention_heads (`int`, *optional*, defaults to 32) : Number of attention heads for each attention layer in the Transformer decoder. num_key_value_heads (`int`, *optional*, defaults to 8) : 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` or `function`, *optional*, defaults to `"silu"`) : The non-linear activation function (function or string) in the decoder. max_position_embeddings (`int`, *optional*, defaults to 9216) : The maximum sequence length that this model might ever be used with. Emu supports up to 9216 tokens, rms_norm_eps (`float`, *optional*, defaults to 1e-05) : 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`. pad_token_id (`int`, *optional*, defaults to 151643) : Padding token id. bos_token_id (`int`, *optional*, defaults to 151849) : Beginning of stream token id. eos_token_id (`int`, *optional*, defaults to 151850) : End of stream token id. tie_word_embeddings (`bool`, *optional*, defaults to `False`) : Whether to tie weight embeddings rope_theta (`float`, *optional*, defaults to 1000000.0) : The base period of the RoPE embeddings. rope_scaling (`Dict`, *optional*) : Dictionary containing the scaling configuration for the RoPE embeddings. NOTE: if you apply new rope type and you expect the model to work on longer `max_position_embeddings`, we recommend you to update this value accordingly. Expected contents: `rope_type` (`str`): The sub-variant of RoPE to use. Can be one of ['default', 'linear', 'dynamic', 'yarn', 'longrope', 'llama3'], with 'default' being the original RoPE implementation. `factor` (`float`, *optional*): Used with all rope types except 'default'. The scaling factor to apply to the RoPE embeddings. In most scaling types, a `factor` of x will enable the model to handle sequences of length x * original maximum pre-trained length. `original_max_position_embeddings` (`int`, *optional*): Used with 'dynamic', 'longrope' and 'llama3'. The original max position embeddings used during pretraining. `attention_factor` (`float`, *optional*): Used with 'yarn' and 'longrope'. The scaling factor to be applied on the attention computation. If unspecified, it defaults to value recommended by the implementation, using the `factor` field to infer the suggested value. `beta_fast` (`float`, *optional*): Only used with 'yarn'. Parameter to set the boundary for extrapolation (only) in the linear ramp function. If unspecified, it defaults to 32. `beta_slow` (`float`, *optional*): Only used with 'yarn'. Parameter to set the boundary for interpolation (only) in the linear ramp function. If unspecified, it defaults to 1. `short_factor` (`list[float]`, *optional*): Only used with 'longrope'. The scaling factor to be applied to short contexts (>> from transformers import Emu3Processor, Emu3ForConditionalGeneration >>> import torch >>> import requests >>> from PIL import Image >>> model = Emu3ForCausalLM.from_pretrained("BAAI/Emu3-Chat-hf", dtype=torch.bfloat16) >>> processor = Emu3Processor.from_pretrained("BAAI/Emu3-Chat-hf") >>> inputs = processor(text=["Can you write me a poem about winter."], return_tensors="pt").to(model.device) >>> generated_ids = model.generate(**inputs, max_new_tokens=100, do_sample=False) >>> processor.batch_decode(generated_ids, skip_special_tokens=True)[0] ``` **Parameters:** config ([Emu3ForCausalLM](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3ForCausalLM)) : 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/v4.57.1/en/main_classes/model#transformers.PreTrainedModel.from_pretrained) method to load the model weights. **Returns:** `[transformers.modeling_outputs.CausalLMOutputWithPast](/docs/transformers/v4.57.1/en/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithPast) or `tuple(torch.FloatTensor)`` A [transformers.modeling_outputs.CausalLMOutputWithPast](/docs/transformers/v4.57.1/en/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithPast) 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 ([Emu3Config](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3Config)) and inputs. - **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/v4.57.1/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. ## Emu3ForConditionalGeneration[[transformers.Emu3ForConditionalGeneration]] #### transformers.Emu3ForConditionalGeneration[[transformers.Emu3ForConditionalGeneration]] [Source](https://github.com/huggingface/transformers/blob/v4.57.1/src/transformers/models/emu3/modeling_emu3.py#L1460) forwardtransformers.Emu3ForConditionalGeneration.forwardhttps://github.com/huggingface/transformers/blob/v4.57.1/src/transformers/models/emu3/modeling_emu3.py#L1507[{"name": "input_ids", "val": ": typing.Optional[torch.LongTensor] = None"}, {"name": "pixel_values", "val": ": typing.Optional[torch.FloatTensor] = None"}, {"name": "image_sizes", "val": ": typing.Optional[torch.Tensor] = None"}, {"name": "attention_mask", "val": ": typing.Optional[torch.Tensor] = None"}, {"name": "position_ids", "val": ": typing.Optional[torch.LongTensor] = None"}, {"name": "past_key_values", "val": ": typing.Optional[transformers.cache_utils.Cache] = None"}, {"name": "inputs_embeds", "val": ": typing.Optional[torch.FloatTensor] = None"}, {"name": "use_cache", "val": ": typing.Optional[bool] = None"}, {"name": "cache_position", "val": ": typing.Optional[torch.LongTensor] = None"}, {"name": "labels", "val": ": typing.Optional[torch.LongTensor] = None"}, {"name": "logits_to_keep", "val": ": typing.Union[int, torch.Tensor] = 0"}, {"name": "**kwargs", "val": ": typing_extensions.Unpack[transformers.utils.generic.TransformersKwargs]"}]- **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/v4.57.1/en/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v4.57.1/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v4.57.1/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) - **pixel_values** (`torch.FloatTensor` 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 ([Emu3Processor](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3Processor) uses `image_processor_class` for processing images). - **image_sizes** (`torch.LongTensor` of shape `(batch_size, 2)`) -- The sizes of the images in the batch, being (height, width) for each image. Image sizes can be obtained using [AutoImageProcessor](/docs/transformers/v4.57.1/en/model_doc/auto#transformers.AutoImageProcessor). See [Emu3ImageProcessor.__call__()](/docs/transformers/v4.57.1/en/model_doc/perceiver#transformers.PerceiverFeatureExtractor.__call__) for details ([]`Emu3Processor`] uses [Emu3ImageProcessor](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3ImageProcessor) for processing images). - **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/v4.57.1/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/v4.57.1/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`). - **cache_position** (`torch.LongTensor` of shape `(sequence_length)`, *optional*) -- Indices depicting the position of the input sequence tokens in the sequence. Contrarily to `position_ids`, this tensor is not affected by padding. It is used to update the cache in the correct position and to infer the complete sequence length. - **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]`. - **logits_to_keep** (`Union[int, torch.Tensor]`, 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).0[transformers.modeling_outputs.CausalLMOutputWithPast](/docs/transformers/v4.57.1/en/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithPast) or `tuple(torch.FloatTensor)`A [transformers.modeling_outputs.CausalLMOutputWithPast](/docs/transformers/v4.57.1/en/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithPast) 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 ([Emu3Config](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3Config)) and inputs. - **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/v4.57.1/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. The [Emu3ForConditionalGeneration](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3ForConditionalGeneration) 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. Example: ```python >>> from transformers import Emu3Processor, Emu3ForConditionalGeneration >>> import torch >>> import requests >>> from PIL import Image >>> model = Emu3ForConditionalGeneration.from_pretrained("BAAI/Emu3-Chat-hf", dtype=torch.bfloat16) >>> processor = Emu3Processor.from_pretrained("BAAI/Emu3-Chat-hf") >>> conversation = [ ... { ... "role": "system", ... "content": [ ... {"type": "text", "text": "You are a helpful assistant."}, ... ], ... }, ... { ... "role": "user", ... "content": [ ... {"type": "image"}, ... {"type": "text", "text": "Please describe the image."}, ... ], ... }, ... ] >>> prompt = processor.apply_chat_template(conversation, add_generation_prompt=True) >>> image = Image.open(requests.get("https://www.ilankelman.org/stopsigns/australia.jpg", stream=True).raw) >>> inputs = processor(images=[image], text=[prompt], return_tensors="pt").to(model.device, torch.bfloat16) >>> generated_ids = model.generate(**inputs, max_new_tokens=100, do_sample=False) >>> processor.batch_decode(generated_ids, skip_special_tokens=True)[0] ``` **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/v4.57.1/en/model_doc/auto#transformers.AutoTokenizer). See [PreTrainedTokenizer.encode()](/docs/transformers/v4.57.1/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.encode) and [PreTrainedTokenizer.__call__()](/docs/transformers/v4.57.1/en/internal/tokenization_utils#transformers.PreTrainedTokenizerBase.__call__) for details. [What are input IDs?](../glossary#input-ids) pixel_values (`torch.FloatTensor` 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 ([Emu3Processor](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3Processor) uses `image_processor_class` for processing images). image_sizes (`torch.LongTensor` of shape `(batch_size, 2)`) : The sizes of the images in the batch, being (height, width) for each image. Image sizes can be obtained using [AutoImageProcessor](/docs/transformers/v4.57.1/en/model_doc/auto#transformers.AutoImageProcessor). See [Emu3ImageProcessor.__call__()](/docs/transformers/v4.57.1/en/model_doc/perceiver#transformers.PerceiverFeatureExtractor.__call__) for details ([]`Emu3Processor`] uses [Emu3ImageProcessor](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3ImageProcessor) for processing images). 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/v4.57.1/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/v4.57.1/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`). cache_position (`torch.LongTensor` of shape `(sequence_length)`, *optional*) : Indices depicting the position of the input sequence tokens in the sequence. Contrarily to `position_ids`, this tensor is not affected by padding. It is used to update the cache in the correct position and to infer the complete sequence length. 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]`. logits_to_keep (`Union[int, torch.Tensor]`, 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:** `[transformers.modeling_outputs.CausalLMOutputWithPast](/docs/transformers/v4.57.1/en/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithPast) or `tuple(torch.FloatTensor)`` A [transformers.modeling_outputs.CausalLMOutputWithPast](/docs/transformers/v4.57.1/en/main_classes/output#transformers.modeling_outputs.CausalLMOutputWithPast) 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 ([Emu3Config](/docs/transformers/v4.57.1/en/model_doc/emu3#transformers.Emu3Config)) and inputs. - **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/v4.57.1/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.