Zj+PjSrSSKrSSKJr SSKJr SSKJr SSK J r SSK J r J r SSKJr SS KJrJrJr SS KJr SS KJr S S KJrJrJrJr S SKJrJrJ r S SK!J"r"J#r#J$r$J%r%J&r&J'r'J(r(J)r)J*r*J+r+ S SK,J-r-J.r.J/r/J0r0J1r1 \"SS9\"SS\555r2\"SS9\"SS\ 555r3\"SS9\"SS\555r4"SS\&5r5"SS\-5r6"SS\5r7"SS \15r8"S!S"\5r9"S#S$\.5r:"S%S&\/5r;"S'S(\5r<"S)S*\$5r="S+S,\05r>"S-S.\#5r?"S/S0\"5r@"S1S2\%5rA"S3S4\$5rB"S5S6\(5rC"S7S8\'5rD"S9S:\)5rE\"S;S<9"S=S>\D55rF"S?S@\D5rG"SASB\\D5rH/SCQrIg)DzPyTorch AltCLIP model.N)strict)initialization)create_bidirectional_mask)BaseModelOutputWithPooling'BaseModelOutputWithPoolingAndProjection)Unpack)TransformersKwargsauto_docstringcan_return_tuple)merge_with_config_defaults)capture_outputs)ChineseCLIPModelChineseCLIPTextAttentionChineseCLIPTextLayerChineseCLIPTextSelfAttention) CLIPConfigCLIPTextConfigCLIPVisionConfig) CLIPMLP CLIPAttention CLIPEncoderCLIPEncoderLayer CLIPOutputCLIPPreTrainedModelCLIPVisionEmbeddingsCLIPVisionModel_get_vector_normimage_text_contrastive_loss)RobertaEmbeddingsRobertaIntermediate RobertaOutput RobertaPoolerRobertaSelfOutputz BAAI/AltCLIP) checkpointc&\rSrSr%SrSr\\S'Sr\\S'Sr \\S'S r \\S 'S r \\S 'S r \ \S'Sr\\-\S'Sr\\-\S'Sr\\S'Sr\\S'Sr\\S'Sr\S-\S'Sr\S-\S'Sr\S-\S'Sr\\S'\"5r\"5rS rg)!AltCLIPTextConfig9a" project_dim (`int`, *optional*, defaults to 768): The dimensions of the teacher model before the mapping layer. Examples: ```python >>> from transformers import AltCLIPTextModel, AltCLIPTextConfig >>> # Initializing a AltCLIPTextConfig with BAAI/AltCLIP style configuration >>> configuration = AltCLIPTextConfig() >>> # Initializing a AltCLIPTextModel (with random weights) from the BAAI/AltCLIP style configuration >>> model = AltCLIPTextModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```i vocab_sizei hidden_sizenum_hidden_layersnum_attention_headsiintermediate_sizegelu hidden_actg?hidden_dropout_probrattention_probs_dropout_probimax_position_embeddingstype_vocab_sizeg{Gz?initializer_factorN pad_token_id bos_token_idr eos_token_id project_dim)__name__ __module__ __qualname____firstlineno____doc__r*int__annotations__r+r-r/r0r2strr3floatr4r5r7r8r9r:r;r=AttributeErrorprojection_dimattention_dropout__static_attributes__r>|/root/GenerationalWealth/GenerationalWealth/venv/lib/python3.13/site-packages/transformers/models/altclip/modular_altclip.pyr(r(9s &JKs!!!s!J'*u*01 #+1#&S&OS $$ L#*  L#*  L#* K#%N&(rLr(c\rSrSrSrSrg)AltCLIPVisionConfigba Example: ```python >>> from transformers import AltCLIPVisionConfig, AltCLIPVisionModel >>> # Initializing a AltCLIPVisionConfig with BAAI/AltCLIP style configuration >>> configuration = AltCLIPVisionConfig() >>> # Initializing a AltCLIPVisionModel (with random weights) from the BAAI/AltCLIP style configuration >>> model = AltCLIPVisionModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ```r>N)r?r@rArBrCrKr>rLrMrOrObs rLrOc(\rSrSr%SrSr\\S'Srg) AltCLIPConfigva Example: ```python >>> from transformers import AltCLIPConfig, AltCLIPModel >>> # Initializing a AltCLIPConfig with BAAI/AltCLIP style configuration >>> configuration = AltCLIPConfig() >>> # Initializing a AltCLIPModel (with random weights) from the BAAI/AltCLIP style configuration >>> model = AltCLIPModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config >>> # We can also initialize a AltCLIPConfig from a AltCLIPTextConfig and a AltCLIPVisionConfig >>> # Initializing a AltCLIPText and AltCLIPVision configuration >>> config_text = AltCLIPTextConfig() >>> config_vision = AltCLIPVisionConfig() >>> config = AltCLIPConfig(text_config=config_text, vision_config=config_vision) ```r<rIr>N) r?r@rArBrCrIrDrErKr>rLrMrRrRvs 0NCrLrRc\rSrSrSrg) AltCLIPOutputr>Nr?r@rArBrKr>rLrMrUrUrLrUc\rSrSrSrg)AltRobertaEmbeddingsr>NrWr>rLrMrZrZrXrLrZc(^\rSrSrU4SjrSrU=r$)AltRobertaSelfAttentionc2>[TU]U5 SUlg)NF)super__init__ is_causalselfconfig __class__s rMra AltRobertaSelfAttention.__init__s  rL)rbr?r@rArBrarK __classcell__rfs@rMr]r]s rLr]c\rSrSrSrg)AltRobertaSelfOutputr>NrWr>rLrMrlrlrXrLrlc(^\rSrSrU4SjrSrU=r$)AltRobertaAttentioncb>[TU]5 [U5Ul[ U5UlgN)r`rar]rdrloutputrcs rMraAltRobertaAttention.__init__s& +F3 *62 rL)rsrdrhrjs@rMroros 33rLroc\rSrSrSrg)AltRobertaIntermediater>NrWr>rLrMrvrvrXrLrvc\rSrSrSrg)AltRobertaOutputr>NrWr>rLrMryryrXrLryc(^\rSrSrU4SjrSrU=r$)AltRobertaLayerc>[TU]5 [U5Ul[ U5Ul[ U5Ulgrr)r`raro attentionrv intermediateryrsrcs rMraAltRobertaLayer.__init__s3 ,V426:&v. rL)rrrsrhrjs@rMr|r|s //rLr|c4^\rSrSrSrS\4U4SjjrSrU=r$)AltRobertaEncoderz Transformer encoder consisting of `config.num_hidden_layers` self attention layers. Each layer is a [`AltRobertaEncoderLayer`]. Args: config: AltCLIPTextConfig rec>[TU]U5 [R"[ UR 5Vs/sHn[ U5PM sn5Ulgs snfrr)r`rann ModuleListranger-r|layers)rdre_rfs rMraAltRobertaEncoder.__init__sD  mmeFLdLdFe$fFe_V%NrWr>rLrMrrrXrLrc\rSrSrSrg)AltCLIPAttentionr>NrWr>rLrMrrrXrLrc\rSrSrSrg) AltCLIPMLPr>NrWr>rLrMrrrXrLrc0^\rSrSrS\4U4SjjrSrU=r$)AltCLIPEncoderLayerrec$>[TU]U5 grr)r`rarcs rMraAltCLIPEncoderLayer.__init__s  rLr>)r?r@rArBrOrarKrirjs@rMrrs!2!!rLrc\rSrSrSrg)AltCLIPEncoderr>NrWr>rLrMrrrXrLrc\rSrSrSrg)AltCLIPVisionEmbeddingsr>NrWr>rLrMrrrXrLrcL\rSrSr\\S.r\R"5S5r Sr g)AltCLIPPreTrainedModel hidden_states attentionsc URRn[U[5(a[R "UR SURS-U-S9 [R "URRURRU-S9 [R "URRURRU-S9 [R"UR[R"UR 5R#S55 g[U[$5(aURS-SURR&-S--U-nURS-U-n[R "UR(RUS9 [R "UR*RUS9 [R "UR,RUS9 [R "UR.RUS9 g[U[05(aURR2S-SURR&-S--U-nSURR2-S-U-n[R "UR4RUS9 [R "UR6RUS9 g[U[85(as[R "UR:RUR<S-U-S9 [R "UR>RUR@S-U-S9 g[U[BRD5(aA[RF"URH5 [RJ"UR5 g[U[BRL5(aO[R "URSUS9 URHb![RF"URH5 gg[U[BRN5(ay[R "URSUS9 URPbK[SURSS 5(d.[RF"URURP5 ggg[U[T5(a|[R"UR[R"URRVS 5R#S55 [RF"URX5 gg) zInitialize the weightsgg)meanstd)r)r6rN_is_hf_initializedFr)-rer8 isinstancerinitnormal_class_embedding embed_dimpatch_embeddingweightinitializer_rangeposition_embeddingcopy_ position_idstorcharange num_positionsexpandrr-q_projk_projv_projout_projrr+fc1fc2 AltCLIPModeltext_projectiontext_embed_dimvisual_projectionvision_embed_dimr LayerNormzeros_biasones_Linear Embedding padding_idxgetattrrZshapetoken_type_ids)rdmodulefactor in_proj_std out_proj_stdfc_stds rM _init_weights$AltCLIPPreTrainedModel._init_weightss// f5 6 6 LL//cv?O?OQU?UX^?^ _ LL//66FMM<[<[^d7j-  4 5 5 JJv**ELL9L9L9R9RSU9V,W,^,^_f,g h KK-- .6rLr>N) r?r@rArBrr_can_record_outputsrno_gradrrKr>rLrMrrs*,&  ]]_+/+/rLrc(^\rSrSrU4SjrSrU=r$)AltCLIPVisionModelic $>[TU]"S0UD6$)a Examples: ```python >>> import httpx >>> from io import BytesIO >>> from PIL import Image >>> from transformers import AutoProcessor, AltCLIPVisionModel >>> model = AltCLIPVisionModel.from_pretrained("BAAI/AltCLIP") >>> processor = AutoProcessor.from_pretrained("BAAI/AltCLIP") >>> url = "http://images.cocodataset.org/val2017/000000039769.jpg" >>> with httpx.stream("GET", url) as response: ... image = Image.open(BytesIO(response.read())) >>> inputs = processor(images=image, return_tensors="pt") >>> outputs = model(**inputs) >>> last_hidden_state = outputs.last_hidden_state >>> pooled_output = outputs.pooler_output # pooled CLS states ```r>)r`forwardrd super_kwargsrfs rMrAltCLIPVisionModel.forwards.w...rLr>)r?r@rArBrrKrirjs@rMrrs //rLraE The model behaves as an encoder following the architecture described in *Attention is all you need*_ by Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser and Illia Polosukhin. .. _*Attention is all you need*: https://huggingface.co/papers/1706.03762 ) custom_introc*^\rSrSr%\\S'SrSr\\ S.r SU4Sjjr \ \ \SS\R S-S \R S-S \R S-S \R S-S \R S-S \\S\\-4Sjj555rSrU=r$)AltRobertaModeli4retextword_embeddingsrc>[TU]U5 [U5Ul[ U5UlU(a [ U5OSUlUR5 g)z^ add_pooling_layer (bool, *optional*, defaults to `True`): Whether to add a pooling layer N) r`rarZ embeddingsrencoderrpooler post_init)rdreadd_pooling_layerrfs rMraAltRobertaModel.__init__GsE  .v6(0 2C&v.  rLN input_idsattention_maskrr inputs_embedskwargsreturnc USLUSL- (a [S5eURUUUUS9n[URUUS9nUR"U4SU0UD6nUSnUR bUR U5OSn [ UU S9$)a Examples: ```python >>> from transformers import AutoTokenizer, AltRobertaModel >>> model = AltRobertaModel.from_pretrained("openai/alt_roberta-vit-base-patch32") >>> tokenizer = AutoTokenizer.from_pretrained("openai/alt_roberta-vit-base-patch32") >>> inputs = tokenizer(["a photo of a cat", "a photo of a dog"], padding=True, return_tensors="pt") >>> outputs = model(**inputs) >>> last_hidden_state = outputs.last_hidden_state >>> pooled_output = outputs.pooler_output # pooled (EOS token) states ```Nz:You must specify exactly one of input_ids or inputs_embeds)rrrr)rerrrr)last_hidden_state pooler_output) ValueErrorrrrerrr) rdrrrrrrencoder_outputssequence_output pooled_outputs rMrAltRobertaModel.forwardSs6  -t"; <YZ Z%)' ( 3;;')  ,,  )   *!,8< 8O O4UY )-'  rL)rrr)TNNNNN)r?r@rArBr(rEinput_modalities_input_embed_layerr|r]rrar rr rTensorr r tuplerrrKrirjs@rMrr4s  *(-  *..2.2,0-1 3 <<$&3  t+3  t+ 3 llT) 3 ||d* 3 +,3  + +3  3 rLrc^\rSrSr%\\S'SrSrSrU4Sjr \ \ SS\ RS-S \ RS-S \ RS-S \ RS-S \ RS-S \\S\\-4Sjj55rSrU=r$)AltCLIPTextModelirerword_embeddingrobertac>[TU]U5 [USS9Ul[R "UR UR5Ul[R"UR URS9Ul UR5 g)NF)r)eps) r`rarrrrr+r=transformationrlayer_norm_epspre_LNrrcs rMraAltCLIPTextModel.__init__se  &vG  ii(:(:F>> from transformers import AutoProcessor, AltCLIPTextModel >>> model = AltCLIPTextModel.from_pretrained("BAAI/AltCLIP") >>> processor = AutoProcessor.from_pretrained("BAAI/AltCLIP") >>> texts = ["it's a cat", "it's a dog"] >>> inputs = processor(text=texts, padding=True, return_tensors="pt") >>> outputs = model(**inputs) >>> last_hidden_state = outputs.last_hidden_state >>> pooled_output = outputs.pooler_output # pooled CLS states ```)rrrrrrN)rrrrr>)rrrrrr) rdrrrrrroutputsrprojection_staters rMrAltCLIPTextModel.forwards:,, ))%'    "!*++o6 ..?(A. 6.'!//))   rL)rrrr)r?r@rArBr(rErrbase_model_prefixrar r rrr r rrrrKrirjs@rMrrs  )!*..2.2,0-1 3 <<$&3  t+3  t+ 3 llT) 3 ||d* 3 +,3  8 83 3 rLrc(^\rSrSr%\\S'S\4U4SjjrU4SjrU4SjrSS\ RS-S\ RS-S \ RS-S \ RS-S \ RS-S \ S-S \ S\\S\\-4SjjrSrU=r$)rirec>[TU]U5 URnURUl[ R URR5Ul[R URR5Ul grr) r`ra text_configr=rr _from_configre text_modelr vision_config vision_model)rdrerrfs rMraAltCLIPModel.__init__sb  (( )55*77 8O8OP.;;DKK[TU]"S0UD6$)a Examples: ```python >>> import torch >>> from transformers import AutoProcessor, AltCLIPModel >>> model = AltCLIPModel.from_pretrained("BAAI/AltCLIP") >>> processor = AutoProcessor.from_pretrained("BAAI/AltCLIP") >>> inputs = processor(text=["a photo of a cat", "a photo of a dog"], padding=True, return_tensors="pt") >>> with torch.inference_mode(): ... text_features = model.get_text_features(**inputs) ```r>)r`get_text_featuresrs rMrAltCLIPModel.get_text_featuressw(8<88rLc $>[TU]"S0UD6$)a Examples: ```python >>> import torch >>> from transformers import AutoProcessor, AltCLIPModel >>> from transformers.image_utils import load_image >>> model = AltCLIPModel.from_pretrained("BAAI/AltCLIP") >>> processor = AutoProcessor.from_pretrained("BAAI/AltCLIP") >>> url = "http://images.cocodataset.org/val2017/000000039769.jpg" >>> image = load_image(url) >>> inputs = processor(images=image, return_tensors="pt") >>> with torch.inference_mode(): ... image_features = model.get_image_features(**inputs) ```r>)r`get_image_featuresrs rMrAltCLIPModel.get_image_featuress&w)9L99rLNr pixel_valuesrrr return_lossinterpolate_pos_encodingrrc 4UR"SUUS.UD6n UR"SUUUUS.UD6n U Sn URU 5n U Sn URU 5n U [ U 5- n U [ U 5- n [ R "XR5RU R55n XRR5RU R5-n U R5nSnU(a [U 5n[UUU U U U U S9$)u return_loss (`bool`, *optional*): Whether or not to return the contrastive loss. Examples: ```python >>> import torch >>> from transformers import AutoProcessor, AltCLIPModel >>> from transformers.image_utils import load_image >>> model = AltCLIPModel.from_pretrained("OFA-Sys/chinese-clip-vit-base-patch16") >>> processor = AutoProcessor.from_pretrained("OFA-Sys/chinese-clip-vit-base-patch16") >>> url = "https://clip-cn-beijing.oss-cn-beijing.aliyuncs.com/pokemon.jpeg" >>> image = load_image(url) >>> inputs = processor(text=["杰尼龟", "妙蛙种子", "小火龙", "皮卡丘"], images=image, return_tensors="pt", padding=True) >>> with torch.inference_mode(): ... outputs = model(**inputs) >>> logits_per_image = outputs.logits_per_image # this is the image-text similarity score >>> probs = logits_per_image.softmax(dim=1) # we can take the softmax to get the label probabilities ```)rr)rrrrr6N)losslogits_per_imagelogits_per_text text_embeds image_embedstext_model_outputvision_model_outputr>)rrrrrrmatmulttodevice logit_scaleexpr rU)rdrrrrrrrrvision_outputs text_outputsr#r"r!r rs rMrAltCLIPModel.forwards?F** %%=    ))%    &a( --l; "1o **;7 $&6|&DD !$4[$AA  ,,{NN4D4G4G HZHZ4[\),<,<,@,@,B,E,EkFXFX,YY*,,. .?D-+#%* .  rL)rrr)NNNNNNF)r?r@rArBrRrErarrr LongTensor FloatTensorrboolr r rrUrrKrirjs@rMrrs W}W9":..215.2.204#').L ##d*L ''$.L  t+ L  t+ L &&- L D[L #'L +,L   L L rLr)rrrrr(rOrR)JrCrtorch.nnrhuggingface_hub.dataclassesrrr masking_utilsrmodeling_outputsrrprocessing_utilsr utilsr r r utils.genericr utils.output_capturingr"chinese_clip.modeling_chinese_cliprrrrclip.configuration_cliprrrclip.modeling_cliprrrrrrrrrr roberta.modeling_robertar!r"r#r$r%r(rOrRrUrZr]rlrorvryr|rrrrrrrrrrrr__all__r>rLrMr@s .&6'II75 SR   >*$)$)+$)N>* * + $>*J+8 J  , :  , 323 0  } /*/ h h }  }   !*!  [  2 2/02/j//6L ,L L ^B -B J| #%;| ~ rL