^jh.XddlZddlmZddlmZmZmZmZmZddlm Z ddl m Z ddl m Z mZddlmZdd lmZmZmZdd lmZd d lmZej2eZGd deZGddeZGddej<ZGddeZ GddeZ!GddeZ"gdZ#y)N)nn)LlavaCausalLMOutputWithPastLlavaForConditionalGeneration LlavaModelLlavaModelOutputWithPastLlavaPreTrainedModel)ACT2FN)Cache)BaseModelOutputWithPastBaseModelOutputWithPooling)Unpack)TransformersKwargsauto_docstringlogging)can_return_tuple)VipLlavaConfigc eZdZy)VipLlavaModelOutputWithPastN__name__ __module__ __qualname__x/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/vipllava/modular_vipllava.pyrr&rrc eZdZy)VipLlavaCausalLMOutputWithPastNrrrrr r *rrr c*eZdZdeffd ZdZxZS)VipLlavaMultiModalProjectorconfigcHt|t|jtrdnt |j}t j||jjz|j|_ t j||jjz|jjd|_t|j |_t j|jj|jjd|_y)Nr)epsT)bias)super__init__ isinstancevision_feature_layersintlenr LayerNorm vision_config hidden_sizeprojector_layernorm_epsprojector_layernormLinear text_configlinear_1r projector_hidden_actactlinear_2)selfr#num_feature_layers __class__s rr(z$VipLlavaMultiModalProjector.__init__/s ",V-I-I3"OQUXY_YuYuUv#%<< !5!5!A!A AvGeGe$   !5!5!A!A A    * *  &556 &"4"4"@"@&BTBTB`B`gkl rc|j|}|j|}|j|}|j|}|SN)r1r4r6r7)r8 hidden_statess rforwardz#VipLlavaMultiModalProjector.forward>sB00?  m4 /  m4 r)rrrrr(r> __classcell__)r:s@rr"r".s m~ mrr"c eZdZy)VipLlavaPreTrainedModelNrrrrrArAFrrrAc~eZdZeed ddej deeezdzde e de e zfdZ ee dd ejdzdej dzd ejdzd ejdzd edzd ej dzdeeezdzdedzde e de ezfdZy) VipLlavaModelzWObtains image last hidden states from the vision tower and apply multimodal projection.) custom_introN pixel_valuesr*kwargsreturnc n||n|jj}d|d<|j|fi|}t|tr|j |ddddf}n<|Dcgc]}|j |ddddf}}t j|d}|j|}||_ |Scc}w)\ pixel_values (`torch.FloatTensor]` of shape `(batch_size, channels, height, width)`): The tensors corresponding to the input images. vision_feature_layers (`Union[int, list[int]]`, *optional*): The vision feature layer, or the list of indexes of the layers to select the vision feature. NToutput_hidden_statesr)dim) r#r* vision_towerr)r+r=torchcatmulti_modal_projector pooler_output)r8rEr*rF image_outputsimage_featuresindexs rget_image_featuresz VipLlavaModel.get_image_featuresKs$&;%F !DKKLmLm *.%&)))    +S 1*889NOPQSTSUPUVNVkkEm99%@ABGkNk"YY~2>N33NC&4 # ls B2 input_idsattention_mask position_idspast_key_values inputs_embeds use_cache lm_kwargsc ||n|jj}|du|duz r td||j|}|i|j ||j } | j |j|j} |j||| } |j| | }|jd|||||d| } t| j| j| j| j | nd} | S)z vision_feature_layers (`Union[int, list[int]]`, *optional*): The vision feature layer, or the list of indexes of the layers to select the vision feature. Nz:You must specify exactly one of input_ids or inputs_embedsrEr*)rZrS)rWrXrYrZr[)last_hidden_staterYr= attentionsimage_hidden_statesr)r#r* ValueErrorget_input_embeddingsrUrQtodevicedtypeget_placeholder_maskmasked_scatterlanguage_modelrr_rYr=r`)r8rVrErWrXrYrZr*r[r\rSspecial_image_maskoutputsoutputs rr>zVipLlavaModel.forwardssF(&;%F !DKKLmLm   -t"; <YZ Z  7D557 BM  #!44)AV5m ,..}/C/C]EXEXYN!%!:!:~";" *889K^\M+>4+>+>, )%+' ,  , -%77#33!//))2>2JPT   rr<)NNNNNNNN)rrrrrrN FloatTensorr+listrrtupler rU LongTensorTensorr boolrr>rrrrCrCJs]n 9="''" #T#Y5"+, " + + ""H.215.204(,268>> import torch >>> from PIL import Image >>> import httpx >>> from io import BytesIO >>> from transformers import AutoProcessor, VipLlavaForConditionalGeneration >>> model = VipLlavaForConditionalGeneration.from_pretrained("llava-hf/vip-llava-7b-hf", device_map="auto", dtype=torch.float16) >>> processor = AutoProcessor.from_pretrained("llava-hf/vip-llava-7b-hf") >>> prompt = "A chat between a curious human and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the human's questions.###Human: \n{}###Assistant:" >>> question = "Can you please describe this image?" >>> prompt = prompt.format(question) >>> url = "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/diffusers/compel-neg.png" >>> with httpx.stream("GET", url) as response: ... image = Image.open(BytesIO(response.read())) >>> inputs = processor(text=text, images=image, return_tensors="pt").to(0, torch.float16) >>> # Generate >>> generate_ids = model.generate(**inputs, max_new_tokens=20) >>> processor.decode(generate_ids[0][len(inputs["input_ids"][0]):], skip_special_tokens=True) The image features a brown and white cat sitting on a green surface, with a red ball in its ```N)rVrErWrXrYrZr[r*)logitsrw vocab_size)lossrzrYr=r`rar)r#r*rvr_r)r+slicelm_head loss_functionr3r{r rYr=r`ra)r8rVrErWrXrYrZr*rwr[rxr\rkr= slice_indicesrzr|s rr>z(VipLlavaForConditionalGeneration.forwardsj&;%F !DKKLmLm 0:tzz 0 %)%+'"7 0  0  11 8B>SV8W~ot4]k mA}a,?@A  %%VFt{{OfOfOqOq%rD-#33!//)) ' ; ;   rr<) NNNNNNNNNr)rrrrrNrmr+rnrrror rUrrprqr rrr r>rrrrtrts9= ''  #T#Y5 +,  + +   ".215.204(,268<*.!%-.R ##d*R ''$.R  t+ R &&- R  R ((4/R  #T#Y5R   4'R $;R ell*R ./R  / /R R rrt)rCrtrA)$rNr(transformers.models.llava.modeling_llavarrrrr activationsr cache_utilsr modeling_outputsr r processing_utilsrutilsrrr utils.genericrconfiguration_vipllavar get_loggerrloggerrr Moduler"rArCrt__all__rrrrs " S&@@-2   H % ":  %@ "))0 2 `J`Fg 'Dg T [r