sg~dZddlmZddlmZmZmZmZddlZddl Zddlm Z ddl m Z ddl mZdd lmZdd lmZdd lmZmZmZmZd d lmZmZddlmZej:eZdZ dZ!eGddeZ"Gdde jFZ$dZ%ede%GddeZ&dZ'ede%Gdde&eZ(y)zPyTorch Llava model.) dataclass)ListOptionalTupleUnionN)nn)ACT2FN)GenerationMixin) ModelOutput)PreTrainedModel)add_start_docstrings%add_start_docstrings_to_model_forwardloggingreplace_return_docstrings) AutoModelAutoModelForCausalLM) LlavaConfigrzllava-hf/llava-1.5-7b-hfceZdZUdZdZeejed<dZ ejed<dZ ee ejed<dZ ee ejed<dZee ejed<dZeejed<y) LlavaCausalLMOutputWithPasta Base class for Llava causal language model (or autoregressive) outputs. Args: 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 (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`): Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape `(batch_size, num_heads, sequence_length, embed_size_per_head)`) 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. image_hidden_states (`torch.FloatTensor`, *optional*): A `torch.FloatTensor` of size (batch_size, num_images, sequence_length, hidden_size)`. image_hidden_states of the model produced by the vision encoder and after projecting the last hidden state. Nlosslogitspast_key_values hidden_states attentionsimage_hidden_states)__name__ __module__ __qualname____doc__rrtorch FloatTensor__annotations__rrrrrrr[/var/www/html/venv/lib/python3.12/site-packages/transformers/models/llava/modeling_llava.pyrr.s<)-D(5$$ %, $FE  $9=OXd5#4#456=8 __classcell__r:s@r(r*r*Vsm{mr'r*ac This model inherits from [`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. Parameters: config ([`LlavaConfig`] or [`LlavaVisionConfig`]): 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 [`~PreTrainedModel.from_pretrained`] method to load the model weights. zSThe bare LLaMA Model outputting raw hidden-states without any specific head on top.c4eZdZeZdZdZdgZdZdZ dZ dZ dZ y)LlavaPreTrainedModelmodelTLlavaVisionAttentionrct|jdr|jjn|jjj}t|dr'|jj j d|t|tjtjfrY|jj j d||j%|jj jyyt|tjrf|jj j d||j2|jj |jjyyy)Ninitializer_rangeclass_embeddingg)meanstd)hasattrr+rFr3rGdatanormal_ isinstancerr0Conv2dweightr-zero_ Embedding padding_idx)r9modulerIs r( _init_weightsz"LlavaPreTrainedModel._init_weightss! t{{$78 KK ) )((:: 6, -  " " ' ' / /Sc / B fryy"))4 5 MM   & &CS & 9{{&   &&('  - MM   & &CS & 9!!- ""6#5#56<<>..r'N) rr r!r config_classbase_model_prefixsupports_gradient_checkpointing_no_split_modules_skip_keys_device_placement_supports_cache_class_supports_flash_attn_2_supports_sdparTr&r'r(rBrBvs: L&*#/0"3 !N?r'rBa Args: input_ids (`torch.LongTensor` of shape `(batch_size, sequence_length)`): Indices of input sequence tokens in the vocabulary. Padding will be ignored by default should you provide it. Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and [`PreTrainedTokenizer.__call__`] for details. [What are input IDs?](../glossary#input-ids) pixel_values (`torch.FloatTensor` of shape `(batch_size, num_channels, image_size, image_size)): The tensors corresponding to the input images. Pixel values can be obtained using [`AutoImageProcessor`]. See [`CLIPImageProcessor.__call__`] for details ([]`LlavaProcessor`] uses [`CLIPImageProcessor`] 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) Indices can be obtained using [`AutoTokenizer`]. See [`PreTrainedTokenizer.encode`] and [`PreTrainedTokenizer.__call__`] for details. If `past_key_values` is used, optionally only the last `decoder_input_ids` have to be input (see `past_key_values`). If you want to change padding behavior, you should read [`modeling_opt._prepare_decoder_attention_mask`] and modify to your needs. See diagram 1 in [the paper](https://arxiv.org/abs/1910.13461) for more information on the default strategy. - 1 indicates the head is **not masked**, - 0 indicates the head is **masked**. 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 (`tuple(tuple(torch.FloatTensor))`, *optional*, returned when `use_cache=True` is passed or when `config.use_cache=True`): Tuple of `tuple(torch.FloatTensor)` of length `config.n_layers`, with each tuple having 2 tensors of shape `(batch_size, num_heads, sequence_length, embed_size_per_head)`) and 2 additional tensors of shape `(batch_size, num_heads, encoder_sequence_length, embed_size_per_head)`. Contains pre-computed hidden-states (key and values in the self-attention blocks and in the cross-attention blocks) that can be used (see `past_key_values` input) to speed up sequential decoding. If `past_key_values` are used, the user can optionally input only the last `decoder_input_ids` (those that don't have their past key value states given to this model) of shape `(batch_size, 1)` instead of all `decoder_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. vision_feature_layer (`int`, *optional*, defaults to -2): The index of the layer to select the vision feature. vision_feature_select_strategy (`str`, *optional*, defaults to `"default"`): The feature selection strategy used to select the vision feature from the vision backbone. Can be one of `"default"` or `"full"`. 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_attentions (`bool`, *optional*): Whether or not to return the attentions tensors of all attention layers. See `attentions` under returned tensors for more detail. output_hidden_states (`bool`, *optional*): Whether or not to return the hidden states of all layers. See `hidden_states` under returned tensors for more detail. return_dict (`bool`, *optional*): Whether or not to return a [`~utils.ModelOutput`] instead of a plain tuple. 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. zIThe LLAVA model which consists of a vision backbone and a language model.c%LeZdZdeffd ZdZdZdZdZdZ dZ d Z d!d e e d ejfd Zd ej$de defdZdZeeeee d"dej6d ej$de ej8de ej6de eej$de ej$de e de ede ej6de ede ede ede ede ej6de d ee eff dZ! d#d Z"xZ#S)$LlavaForConditionalGenerationr+ct||tj|j|_t ||_|jj|_ tj|j|_ |jj|jjnd|_ |jy)N)r.r/r from_configr1 vision_towerr*multi_modal_projectorr3 vocab_sizerlanguage_modelr+ pad_token_id post_initr8s r(r/z&LlavaForConditionalGeneration.__init__s  %11&2F2FG%=f%E" ,,772>>v?Q?QR8< 8P8P8\DKK44bd r'c6|jjSr<)reget_input_embeddingsr9s r(riz2LlavaForConditionalGeneration.get_input_embeddingss""7799r'c:|jj|yr<)reset_input_embeddings)r9values r(rlz2LlavaForConditionalGeneration.set_input_embeddingss 007r'c6|jjSr<)reget_output_embeddingsrjs r(roz3LlavaForConditionalGeneration.get_output_embeddingss""88::r'c:|jj|yr<)reset_output_embeddings)r9new_embeddingss r(rqz3LlavaForConditionalGeneration.set_output_embeddingss 11.Ar'c:|jj|yr<)re set_decoder)r9decoders r(rtz)LlavaForConditionalGeneration.set_decoders ''0r'c6|jjSr<)re get_decoderrjs r(rwz)LlavaForConditionalGeneration.get_decoder""..00r'c6|jjSr<)re tie_weightsrjs r(rzz)LlavaForConditionalGeneration.tie_weightsrxr'new_num_tokensreturnc|jj||}|j|jj_|j|_|Sr<)reresize_token_embeddingsnum_embeddingsr+r3rd)r9r{pad_to_multiple_of model_embedss r(r~z5LlavaForConditionalGeneration.resize_token_embeddings sF**BB>Sef -9-H-H *&55r' pixel_valuesvision_feature_layervision_feature_select_strategyc|j|d}|j|}|dk(r |ddddf}n*|dk(r|}n"td|jj|j |}|S)a Obtains image last hidden states from the vision tower and apply multimodal projection. Args: pixel_values (`torch.FloatTensor]` of shape `(batch_size, channels, height, width)`) The tensors corresponding to the input images. vision_feature_layer (`int`): The index of the layer to select the vision feature. vision_feature_select_strategy (`str`): The feature selection strategy used to select the vision feature from the vision backbone. Can be one of `"default"` or `"full"` Returns: image_features (`torch.Tensor`): Image feature tensor of shape `(num_images, image_length, embed_dim)`). T)output_hidden_statesdefaultNrfullz$Unexpected select feature strategy: )rbr ValueErrorr+rrc)r9rrr image_outputsselected_image_featurer=s r(get_image_featuresz0LlavaForConditionalGeneration.get_image_featuress")),T)R !.!q"u+EE  <4#8 8 / ?P0PQ ++  y 8K8KTaThTh  %{{  ^-A-A-J^J^    ::]+T[[-E-EY__eneueuL &,,   ] +   /   /+<( (**=9=J-YjJj  =CMSdDd=eL(99 :#ZZ  'UZZ H\H\ @E=*;;<  "4";";B"?!"C|TUW[T[G\G_G_`mGn"n n ??#5UZZHOOPRSVWWD#6q"#v#>#A#A-#PPL , .   ! ! #~';';CR'@'F'F'H H[\a\e\ef~\\AA>>H\JGH  /=.G.G.I.Q.QRTV_.`.c.cdq.r*+ 22,33B7!;IIK_cdKdghi &+[[d>O>O1O%P" {-m[.HI:; 67 >L 4lLPPr') output_typerUrrrrrr use_cacheoutput_attentionsr return_dictcache_positionnum_logits_to_keepc > | | n|jj} | | n|jj} | | n|jj} ||n|jj}||n|jj }|du|duz r t d| | t dd}|}|j|}||jjk(jdj|jjkxs|jddk(xr|du}d}||j|||}|rtjd|jddk7rJ|j!||||| \}}} }t#j$|jd|j& }n>|d d ddddddd f}t#j(|j+jd d k(\}}|jd}|jd}t#j,|jd |f|j.|j& }||j1dk}||}||}d |||f<t#j2||dd| dffd }t#j|d j5ddz }t#j$|jd|j& | d}n|||jjk(jj7}|jd |jdz}||k7rt d|d|||jjk(j5dj9|j;|j&}|j;|j&|j.}|j=||}|j?||||| | | | || }|d }d} | <||dd|jddz dfj;|j&}!|dddddf|!j;|j&d k7jA}"| dddf|!j;| j&d k7jA}#n1|dddddfjA}"| dddfjA}#tCjD}$|$|"jGd|"j1d|#jGdj;|"j&} | s|f|ddz}%| | f|%zS|%StI| ||jJ|jL|jN||SdS)ah Args: 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]`. num_logits_to_keep (`int`, *optional*): Calculate logits for the last `num_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. Returns: Example: ```python >>> from PIL import Image >>> import requests >>> from transformers import AutoProcessor, LlavaForConditionalGeneration >>> model = LlavaForConditionalGeneration.from_pretrained("llava-hf/llava-1.5-7b-hf") >>> processor = AutoProcessor.from_pretrained("llava-hf/llava-1.5-7b-hf") >>> prompt = "USER: \nWhat's the content of the image? ASSISTANT:" >>> url = "https://www.ilankelman.org/stopsigns/australia.jpg" >>> image = Image.open(requests.get(url, stream=True).raw) >>> inputs = processor(images=image, text=prompt, return_tensors="pt") >>> # Generate >>> generate_ids = model.generate(**inputs, max_new_tokens=15) >>> processor.batch_decode(generate_ids, skip_special_tokens=True, clean_up_tokenization_spaces=False)[0] "USER: \nWhat's the content of the image? ASSISTANT: The image features a busy city street with a stop sign prominently displayed" ```Nz:You must specify exactly one of input_ids or inputs_embedszdYou cannot specify both pixel_values and inputs_embeds at the same time, and must specify either oneFrr`)rrraExpanding inputs for image tokens in LLaVa should be done in processing. Please add `patch_size` and `vision_feature_select_strategy` to the model's processing config or set directly with `processor.patch_size = {{patch_size}}` and processor.vision_feature_select_strategy = {{vision_feature_select_strategy}}`. Using processors without these attributes in the config is deprecated and will throw an error in v4.50.)rrrrz6Image features and image tokens do not match: tokens: z , features ) rrrrrrrrrr.)rrrrrr)(r+rruse_return_dictrrrrirrrimage_seq_lengthrrlogger warning_oncerr#arangerrfloatonesrsizecat unsqueezeitem expand_asrmasked_scatterrerrCrossEntropyLossviewrrrr)&r9rrrrrrrrrrrrrrrlegacy_processingr=first_layer_past_key_value batch_indexnon_attended_tokens target_length past_lengthextended_attention_mask valid_indicesnew_batch_indexnew_non_attended_tokensn_image_tokensn_image_featuresspecial_image_maskoutputsrrshift_attention_mask shift_logits shift_labelsloss_fctoutputs& r(r>z%LlavaForConditionalGeneration.forwardsr2C1N-TXT_T_TqTq$8$D $++JjJj &1%>rB */**#))!,k:(..)00+'!46M6R6RSU6V V "-m"<*=m*L'UV'9P(PQ!&,C^TUXeWeWfTfEg+hno!p$yyQ?II"MPQQ !&n.B.B1.EnNcNc!dfseset!u ''4;;+H+HHMMOTTVN-33A69M9Ma9PP !11 L^L\\ghxgyzdkk;;;2=)M(()  ,..}/C/C]EXEXYN)889K^\M%%)%+'/!5#)1&    )(6a6<<?Q;N9O9Q6Q'R'U'UV\VcVc'd$%c3B3k23G3J3J6==3Y]^3^_jjl %c12g/C/F/Fv}}/UYZ/Z[ffh %c3B3k2==? %c12g99; **,H!!"l&7&7&;O>OPR>S>V>VWcWjWj>kDY,F'+'7D7V# CV C*#33!//))2>2J   QU   r'c f|jj|f|||||d|} |ddk(r|| d<| S)N)rrrrrrr)reprepare_inputs_for_generation) r9rrrrrrrkwargs model_inputss r(rz;LlavaForConditionalGeneration.prepare_inputs_for_generationPs_It**HH  +'))1    !  !,8L (r')NN)NNNNNNNNNNNNNNr)NNNNNN)$rr r!rr/rirlrorqrtrwrzrintrrQr~r#r$strrrrLLAVA_INPUTS_DOCSTRINGrr_CONFIG_FOR_DOC LongTensorTensorrrrrr>rr?r@s@r(r^r^s {:8;B111hsmhjhtht!--EHjm:QQf++AB+FUde'+*.1537=A59.28<-1$(,0/3&*59"#!K ##K ''K !. K u//0 K "$u'8'8"9: K   1 12K 'smK )1 K ))*K D>K $D>K 'tnK d^K !!1!12K  !K " u11 2#K fCK `r'r^))r" dataclassesrtypingrrrrr#torch.utils.checkpointr activationsr generationr modeling_outputsr modeling_utilsr utilsrrrrautorrconfiguration_llavar get_loggerrrr_CHECKPOINT_FOR_DOCrModuler*LLAVA_START_DOCSTRINGrBrr^r&r'r(rs!// !)+- 3,   H %1 $<+$< $