+ jRt^RIHt^RIHt^RIHt^RIHtH t ] P!] 4t ]!RR7]!RR]444t ]!RR7]!R R ]444tR R.tR #) zchameleon model configuration)strict)PreTrainedConfig)RopeParameters)auto_docstringloggingzfacebook/chameleon-7b) checkpointcla]tRt^toRtRtRtRtRtRt Rt Rt ^t ^t Rt^tRtR tR tR tV3R ltR tVtR#)ChameleonVQVAEConfiga| resolution (`int`, *optional*, defaults to 512): Resolution of the input images. base_channels (`int`, *optional*, defaults to 128): Base channel count. channel_multiplier (`list[int]`, *optional*, defaults to `[1, 1, 2, 2, 4]`): Channel multipliers for each resolution. num_res_blocks (`int`, *optional*, defaults to 2): Number of residual blocks. attn_resolutions (`list[int]`, *optional*): Resolutions to apply attention. dropout (`float`, *optional*, defaults to 0.0): Dropout rate. attn_type (`str`, *optional*, defaults to `"vanilla"`): Attention type used in VQ-GAN encoder. Can be "vanilla" or None chameleon_vqgan vq_configi FiNvanilla{Gz?c <V^8dQh/S[;R&S[;R&S[;R&S[;R&S[;R&S[;R&S[;R&S[S[,S[S[R3,,;R &S[;R &S[S[,R ,;R &S[S[,;R &S[;R&#) embed_dimnum_embeddings double_latentlatent_channels resolution in_channels base_channels.channel_multipliernum_res_blocksNattn_resolutionsdropout attn_type)intboollisttuplefloatstr)format __classdict__s"v/home/luca/screen-checking/.venv/lib/python3.14/site-packages/transformers/models/chameleon/configuration_chameleon.py __annotate__!ChameleonVQVAEConfig.__annotate__s./0123456789:;<S E#s(O3E=>?@3i$&-ABS[CDE)r+rr)__name__ __module__ __qualname____firstlineno____doc__ model_typebase_config_keyrrrrrrrrrrrrinitializer_range__annotate_func____static_attributes____classdictcell__)r%s@r&r r sb"#J!OINMOJKM6EN)-GIGr)r caa]tRt^@toRtRtR]/tR.tRt Rt Rt ^ t ^ t ^ tRtRtR tR tR tR t^t^tR tR tR tRt^tR tR tR tR tV3Rlt V3Rlt!Rt"Vt#V;t$#)ChameleonConfigaa model_parallel_size (`int`, *optional*, defaults to 1): Number of shards used when training the model. This will be used in qk layernorm because the original Chameleon inference doesn't do reduction in those layers and each rank has its own biases. swin_norm (`bool`, *optional*, defaults to `False`): Use Swin Transformer normalization. vocabulary_map (`dict`, *optional*): A dictionary containing the vocabulary map from the tokenizer. Used to obtain tokens from the image inputs. ```python >>> from transformers import ChameleonModel, ChameleonConfig >>> # Initializing a chameleon chameleon-7b style configuration >>> configuration = ChameleonConfig() >>> # Initializing a model from the chameleon-7b style configuration >>> model = ChameleonModel(configuration) >>> # Accessing the model configuration >>> configuration = model.config ``` chameleonr past_key_valuesiii+silurgh㈵>TNFr c b<VPf&\PR4\4VnM:\ VP\ 4'd\R/VPBVnVP eVP PR4MRVn\SV`(!R/VBR#)NzJvq_config is None. initializing the ChameleonVQConfig with default values.zr*) r loggerinfor isinstancedictvocabulary_mapgetimage_token_idsuper __post_init__)selfkwargs __class__s&,r&rFChameleonConfig.__post_init__vs >> ! KKd e13DN  - -1CDNNCDNDHDWDWDcd1155i@im ''r)c<V^8dQh/S[;R&S[;R&S[;R&S[;R&S[;R&S[R,;R&S[;R&S[;R &S[;R &S[;R &S[;R &S[R,;R &S[R,;R&S[S[S[,,R,;R&S[;R&S[S[,R,;R&S[R,;R&S[S[,R,;R&S[R,;R&S[R,;R&S[S[,R,;R&S[R,;R&S[;R&#)r vocab_size hidden_sizeintermediate_sizenum_hidden_layersnum_attention_headsNnum_key_value_heads hidden_actmax_position_embeddingsr4 rms_norm_eps use_cache pad_token_id bos_token_id eos_token_idtie_word_embeddingsrope_parametersattention_biasattention_dropoutmodel_parallel_size swin_normr rBmlp_bias)rr#r"rr rrAr)r$r%s"r&r'ChameleonConfig.__annotate__@s<=>?@"ABCD!EFt(GHIJ!'KL#MNOPQR*#ST* UVS /D(,WX%YZ$d*T18[\4K']^s{T)/_`t'abd{"cd&&-4ef4K&ghir))rDr )%r-r.r/r0r1r2r sub_configskeys_to_ignore_at_inferencerLrMrNrOrPrQrRrSr4rTrUrVrWrXrYrZr[r\r]r^r rBr_rFr5r6r7 __classcell__)rIr%s@@r&r9r9@s.J 45K#4"5JK"!&(J#'#LI#L L+,L %48O"'N,/&'"I04I"&NH (mr)r9N)r1huggingface_hub.dataclassesrconfiguration_utilsrmodeling_rope_utilsrutilsrr get_loggerr-r>r r9__all__r*r)r&rjs$.31,   H %23!+!4!H23=(&=(4=(@ 4 5r)