^j> d Z ddlmZ ddlmZ ddlmZ ed e G d d e Z ed e G d
d e Z ed e G d d
e Z g dZ
y)zIdefics model configuration )strict )PreTrainedConfig)auto_docstringzHuggingFaceM4/idefics-9b)
checkpointc e Zd ZU dZddiZdZeed< dZee e z e
eef z ed< dZ eed< d Z ee e z e
eef z ed
< d Z
eed < d
Zeed< dZeed< dZeed< dZeed< dZeez ed< dZeed< dZeed< y) IdeficsVisionConfigidefics_vision hidden_size embed_dimi
image_sizei intermediate_size
patch_size num_hidden_layers num_attention_headsr num_channelsgelu
hidden_actgh㈵>layer_norm_eps attention_dropout{Gz?initializer_rangeg ?initializer_factorN)__name__
__module__ __qualname__
model_type
attribute_mapr int__annotations__r listtupler r r r r r strr floatr r r |/var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/idefics/configuration_idefics.pyr r s "J"K0MIs47Jd3i%S/17 !s!46Jd3i%S/16 s !!L# J NE %(us{( #u# ##r+ r c j e Zd ZU dZdZdZeed< dZe ed< dZ
e ed< d Z e ed
< d Z e ed < dZ
eed
< y)IdeficsPerceiverConfigaB
use_resampler (`bool`, *optional*, defaults to `False`):
Whether or not to use the resampler
resampler_n_latents (`int`, *optional*, defaults to 64):
Number of latent embeddings to resample ("compress") the input sequence to (usually < 128).
resampler_depth (`int`, *optional*, defaults to 6):
Depth of the Perceiver Resampler (Transformer w/ cross attention). Should be shallow (< 3).
resampler_n_heads (`int`, *optional*, defaults to 16):
Number of heads in each Transformer block (for multi-headed self-attention).
resampler_head_dim (`int`, *optional*, defaults to 96):
Dimensionality of each head projection in the Transformer block.
qk_layer_norms_perceiver (`bool`, *optional*, defaults to `False`):
Whether or not to use qk layer norms in perceiver
idefics_percieverF
use_resampler@ resampler_n_latents resampler_depthr resampler_n_heads` resampler_head_dimqk_layer_norms_perceiverN)r r r! __doc__r" r0 boolr% r2 r$ r4 r5 r7 r8 r* r+ r, r. r. / sN
%J M4 !! OS s %* d*r+ r. c e Zd ZU dZdZeedZdZe e
d< dZ e e
d< dZ e e
d < d
Z
e e
d < d Ze e
d
< d Ze e
d< dZee z e
d< dZee
d< dZee
d< dZee
d< dZee
d< dZee
d< dZee
d< d Zee
d < dZe d z e
d < d Z e d z e
d!< d"Z e e e z d z e
d#< d$Z ee
d%< d Z e e
d&< d$Z!ee
d'< d Z"ee
d(< d)Z#e e$z e
d*< d$Z%ee
d+< d Z&ee
d,< d)Z'e e$z e
d-< d$Z(ee
d.< d Z)e*e+z d z e
d/< d Z,e*e+z d z e
d0< fd1Z- xZ.S )2
IdeficsConfigah
additional_vocab_size (`int`, *optional*, defaults to 0):
Additional vocabulary size of the model, typically for the special "![]()
" token. Additional vocab tokens
are always trainable whereas regular vocab tokens can be frozen or not.
alpha_initializer (`str`, *optional*, defaults to `"zeros"`):
Initialization type for the alphas.
alphas_initializer_range (`float`, *optional*, defaults to 0.0):
The standard deviation of the truncated_normal_initializer for initializing the alphas in the Gated Cross
Attention.
alpha_type (`str`, *optional*, defaults to `"float"`):
Whether the gating alphas should be vectors or single floats.
cross_layer_interval (`int`, *optional*, default to 1):
Interval for cross attention (from text to image) layers.
qk_layer_norms (`bool`, *optional*, defaults to `False`):
Whether to add layer norm after q and k
freeze_text_layers (`bool`, *optional*, defaults to `True`):
Whether to freeze text layers
freeze_text_module_exceptions (`bool`, *optional*, defaults to `[]`):
Exceptions to freezing text layers when `freeze_text_layers` is `True`
freeze_lm_head (`bool`, *optional*, defaults to `False`):
Whether to freeze lm head
freeze_vision_layers (`bool`, *optional*, defaults to `True`):
Whether to freeze vision layers
freeze_vision_module_exceptions (`bool`, *optional*, defaults to `[]`):
Exceptions to freezing vision layers when `freeze_vision_layers` is `True`
use_resampler (`bool`, *optional*, defaults to `False`):
Whether to use the Resampler
perceiver_config (`IdeficsPerceiverConfig`, *optional*):
Custom perceiver config or dict
Example:
```python
>>> from transformers import IdeficsModel, IdeficsConfig
>>> # Initializing a Idefics idefics-9b style configuration
>>> configuration = IdeficsConfig()
>>> # Initializing a model from the idefics-9b style configuration
>>> model = IdeficsModel(configuration)
>>> # Accessing the model configuration
>>> configuration = model.config
```idefics)perceiver_config
vision_configi }
vocab_sizer additional_vocab_sizei r i + r r r r r dropoutsilur r r zerosalpha_initializeralphas_initializer_ranger)
alpha_typegư> rms_norm_epsT use_cacheN pad_token_id bos_token_id eos_token_idFtie_word_embeddingscross_layer_intervalqk_layer_normsfreeze_text_layersr* freeze_text_module_exceptionsfreeze_lm_headfreeze_vision_layers freeze_vision_module_exceptionsr0 r? r> c d | j t | _ n4t | j t rt di | j | _ | j t | _ n4t | j t rt di | j | _ t
| di | y )Nr* )r> r.
isinstancedictr? r super
__post_init__)selfkwargs __class__s r, r[ zIdeficsConfig.__post_init__ s ($:$ r[
__classcell__)r^ s @r, r< r< K s + Z J'=PcdKJ!"3"K "s" s !! GUS[ J #u# $s$&) e) J L% It L#* L#* +,L#S /D(, %% !#! ND ##24!4%<4 ND !%$%46#TE\6 M4 48M4**T187;d --4; ( (r+ r< )r< r. r N) r9 huggingface_hub.dataclassesr configuration_utilsr utilsr r r. r< __all__r* r+ r, re s & " . 3 # 56$* $ 7$$ 56+ - + 7+4 56Y($ Y( 7Y(x Mr+