o
h @ s d Z ddlZddlZddlZddlZddlmZmZmZ ddl m
Z
ddl m Z ddl
mZ ddlmZ dd lmZmZmZmZmZmZmZmZmZmZm Z dd
l m Z e e Z!G d d d eZ"d
e#e$ de$fddZ%dddZ&ee"j'e"_'e"j'j dure"j'j j(dddde"j'_ dS dS )z'Configuration base class and utilities. N)AnyOptionalUnion)version ) __version__)custom_object_save)load_gguf_checkpoint) CONFIG_NAMEPushToHubMixinadd_model_info_to_auto_map"add_model_info_to_custom_pipelines cached_file copy_func download_urlextract_commit_hash
is_remote_urlis_torch_availablelogging)is_timm_config_dictc s e Zd ZU dZdZeed< dZeed< i Ze ed f ed< dZ
e ed< i Z e eef ed< d Z
ee eef ed
< d Zee eeee f ed < d Zee ed < fd
dZ fddZdd ZedefddZejdd Zede fddZedefddZ e jdefd dZ ed d Z e jd d Z d]d e ee j f d!e fd"d#Z!e"d^d$d%Z#e$ &d_d'e ee j f d(ee ee j f d)e d*e d+ee ee f d,edd fd-d.Z%e$d'e ee j f dee eef e eef f fd/d0Z&e$d'e ee j f dee eef e eef f fd1d2Z'e$d3e eef dd fd4d5Z(e$d6e ee j f dd fd7d8Z)e$d6e ee j f fd9d:Z*d;d< Z+d=d> Z,d?d@ Z-de eef fdAdBZ.de eef fdCdDZ/d`dFe defdGdHZ0d`dIe ee j f dFe fdJdKZ1d3e eef fdLdMZ2dNefdOdPZ3dQe eef dd fdRdSZ4e$dadUdVZ5e"de eef fdWdXZ6de eef fdYdZZ7d]dbd[d\Z8 Z9S )cPretrainedConfiga8%
Base class for all configuration classes. Handles a few parameters common to all models' configurations as well as
methods for loading/downloading/saving configurations.
A configuration file can be loaded and saved to disk. Loading the configuration file and using this file to
initialize a model does **not** load the model weights. It only affects the model's configuration.
Class attributes (overridden by derived classes):
- **model_type** (`str`) -- An identifier for the model type, serialized into the JSON file, and used to recreate
the correct object in [`~transformers.AutoConfig`].
- **is_composition** (`bool`) -- Whether the config class is composed of multiple sub-configs. In this case the
config has to be initialized from two or more configs of type [`~transformers.PretrainedConfig`] like:
[`~transformers.EncoderDecoderConfig`] or [`~RagConfig`].
- **keys_to_ignore_at_inference** (`List[str]`) -- A list of keys to ignore by default when looking at dictionary
outputs of the model during inference.
- **attribute_map** (`Dict[str, str]`) -- A dict that maps model specific attribute names to the standardized
naming of attributes.
- **base_model_tp_plan** (`Dict[str, Any]`) -- A dict that maps sub-modules FQNs of a base model to a tensor
parallel plan applied to the sub-module when `model.tensor_parallel` is called.
- **base_model_pp_plan** (`Dict[str, Tuple[List[str]]]`) -- A dict that maps child-modules of a base model to a
pipeline parallel plan that enables users to place the child-module on the appropriate device.
Common attributes (present in all subclasses):
- **vocab_size** (`int`) -- The number of tokens in the vocabulary, which is also the first dimension of the
embeddings matrix (this attribute may be missing for models that don't have a text modality like ViT).
- **hidden_size** (`int`) -- The hidden size of the model.
- **num_attention_heads** (`int`) -- The number of attention heads used in the multi-head attention layers of the
model.
- **num_hidden_layers** (`int`) -- The number of blocks in the model.
Setting parameters for sequence generation in the model config is deprecated. For backward compatibility, loading
some of them will still be possible, but attempting to overwrite them will throw an exception -- you should set
them in a [~transformers.GenerationConfig]. Check the documentation of [~transformers.GenerationConfig] for more
information about the individual parameters.
Arg:
name_or_path (`str`, *optional*, defaults to `""`):
Store the string that was passed to [`PreTrainedModel.from_pretrained`] or
[`TFPreTrainedModel.from_pretrained`] as `pretrained_model_name_or_path` if the configuration was created
with such a method.
output_hidden_states (`bool`, *optional*, defaults to `False`):
Whether or not the model should return all hidden-states.
output_attentions (`bool`, *optional*, defaults to `False`):
Whether or not the model should returns all attentions.
return_dict (`bool`, *optional*, defaults to `True`):
Whether or not the model should return a [`~transformers.utils.ModelOutput`] instead of a plain tuple.
is_encoder_decoder (`bool`, *optional*, defaults to `False`):
Whether the model is used as an encoder/decoder or not.
is_decoder (`bool`, *optional*, defaults to `False`):
Whether to only use the decoder in an encoder-decoder architecture, otherwise it has no effect on decoder-only or encoder-only architectures.
cross_attention_hidden_size** (`bool`, *optional*):
The hidden size of the cross-attention layer in case the model is used as a decoder in an encoder-decoder
setting and the cross-attention hidden dimension differs from `self.config.hidden_size`.
add_cross_attention (`bool`, *optional*, defaults to `False`):
Whether cross-attention layers should be added to the model. Note, this option is only relevant for models
that can be used as decoder models within the [`EncoderDecoderModel`] class, which consists of all models
in `AUTO_MODELS_FOR_CAUSAL_LM`.
tie_encoder_decoder (`bool`, *optional*, defaults to `False`):
Whether all encoder weights should be tied to their equivalent decoder weights. This requires the encoder
and decoder model to have the exact same parameter names.
prune_heads (`Dict[int, List[int]]`, *optional*, defaults to `{}`):
Pruned heads of the model. The keys are the selected layer indices and the associated values, the list of
heads to prune in said layer.
For instance `{1: [0, 2], 2: [2, 3]}` will prune heads 0 and 2 on layer 1 and heads 2 and 3 on layer 2.
chunk_size_feed_forward (`int`, *optional*, defaults to `0`):
The chunk size of all feed forward layers in the residual attention blocks. A chunk size of `0` means that
the feed forward layer is not chunked. A chunk size of n means that the feed forward layer processes `n` <
sequence_length embeddings at a time. For more information on feed forward chunking, see [How does Feed
Forward Chunking work?](../glossary.html#feed-forward-chunking).
> Parameters for fine-tuning tasks
architectures (`List[str]`, *optional*):
Model architectures that can be used with the model pretrained weights.
finetuning_task (`str`, *optional*):
Name of the task used to fine-tune the model. This can be used when converting from an original (TensorFlow
or PyTorch) checkpoint.
id2label (`Dict[int, str]`, *optional*):
A map from index (for instance prediction index, or target index) to label.
label2id (`Dict[str, int]`, *optional*): A map from label to index for the model.
num_labels (`int`, *optional*):
Number of labels to use in the last layer added to the model, typically for a classification task.
task_specific_params (`Dict[str, Any]`, *optional*):
Additional keyword arguments to store for the current task.
problem_type (`str`, *optional*):
Problem type for `XxxForSequenceClassification` models. Can be one of `"regression"`,
`"single_label_classification"` or `"multi_label_classification"`.
> Parameters linked to the tokenizer
tokenizer_class (`str`, *optional*):
The name of the associated tokenizer class to use (if none is set, will use the tokenizer associated to the
model by default).
prefix (`str`, *optional*):
A specific prompt that should be added at the beginning of each text before calling the model.
bos_token_id (`int`, *optional*): The id of the _beginning-of-stream_ token.
pad_token_id (`int`, *optional*): The id of the _padding_ token.
eos_token_id (`int`, *optional*): The id of the _end-of-stream_ token.
decoder_start_token_id (`int`, *optional*):
If an encoder-decoder model starts decoding with a different token than _bos_, the id of that token.
sep_token_id (`int`, *optional*): The id of the _separation_ token.
> PyTorch specific parameters
torchscript (`bool`, *optional*, defaults to `False`):
Whether or not the model should be used with Torchscript.
tie_word_embeddings (`bool`, *optional*, defaults to `True`):
Whether the model's input and output word embeddings should be tied. Note that this is only relevant if the
model has a output word embedding layer.
torch_dtype (`str`, *optional*):
The `dtype` of the weights. This attribute can be used to initialize the model to a non-default `dtype`
(which is normally `float32`) and thus allow for optimal storage allocation. For example, if the saved
model is `float16`, ideally we want to load it back using the minimal amount of memory needed to load
`float16` weights. Since the config object is stored in plain text, this attribute contains just the
floating type string without the `torch.` prefix. For example, for `torch.float16` ``torch_dtype` is the
`"float16"` string.
This attribute is currently not being used during model loading time, but this may change in the future
versions. But we can already start preparing for the future by saving the dtype with save_pretrained.
> TensorFlow specific parameters
use_bfloat16 (`bool`, *optional*, defaults to `False`):
Whether or not the model should use BFloat16 scalars (only used by some TensorFlow models).
tf_legacy_loss (`bool`, *optional*, defaults to `False`):
Whether the model should use legacy TensorFlow losses. Legacy losses have variable output shapes and may
not be XLA-compatible. This option is here for backward compatibility and will be removed in Transformers
v5.
loss_type (`str`, *optional*):
The type of loss that the model should use. It should be in `LOSS_MAPPING`'s keys, otherwise the loss will
be automatically inferred from the model architecture.
model_typebase_config_key sub_configsFis_composition
attribute_mapNbase_model_tp_planbase_model_pp_plan _auto_classc s2 |t dv rt d| }t || d S Nr )super__getattribute__ __setattr__)selfkeyvalue __class__ t/var/www/html/construction_image-detection-poc/venv/lib/python3.10/site-packages/transformers/configuration_utils.pyr# s z PretrainedConfig.__setattr__c s4 |dkr|t dv rt d| }t |S r )r! r" )r$ r% r' r) r* r" s z!PretrainedConfig.__getattribute__c
K s | dd| _| dd| _| dd| _| dd| _| dd | _| dd| _| d d| _| d
i | _| d d| _ | d d
| _
| dd| _ | dd| _ | dd | _
| dd| _| dd| _| D ]\}}t| || || qo| dd | _| dd | _| dd | _| dd | _| jd urt| jtstd| jd urt| jtstd| dd }|d urt| j|krt d| d| j d | j d d d | j D | _n| dd | _ | jd ur t| jt r t r d
d l }t!|| j| _| d!d | _"| d"d | _#| d#d | _$| d$d | _%| d%d | _&| d&d | _'| d'd | _(| d(d | _)| d)d | _*d*}| j*d ur_| j*|vr_td+| j* d,| d-d d urmt d. t | d/d0| _+| d1d | _,| d2d | _-d| _.| d3d | _/|0d4drt12d5 | D ]*\}}z t| || W q t3y } zt4d6| d7| d8| | d } ~ ww d S )9N return_dictToutput_hidden_statesFoutput_attentions torchscript torch_dtype use_bfloat16tf_legacy_loss pruned_headstie_word_embeddingschunk_size_feed_forwardr is_encoder_decoder
is_decodercross_attention_hidden_sizeadd_cross_attentiontie_encoder_decoder
architecturesfinetuning_taskid2labellabel2idz)Argument label2id should be a dictionary.z)Argument id2label should be a dictionary.
num_labels You passed along `num_labels=(` with an incompatible id to label map: z.. The number of labels will be overwritten to .c S i | ] \}}t ||qS r) int.0r% r& r) r) r*
z-PretrainedConfig.__init__.. tokenizer_classprefix bos_token_id pad_token_id eos_token_id sep_token_iddecoder_start_token_idtask_specific_params problem_type)
regressionsingle_label_classificationmulti_label_classificationzAThe config parameter `problem_type` was not understood: received za but only 'regression', 'single_label_classification' and 'multi_label_classification' are valid.
xla_devicezThe `xla_device` argument has been deprecated in v4.4.0 of Transformers. It is ignored and you can safely remove it from your `config.json` file. name_or_pathr _commit_hashattn_implementationtransformers_versiongradient_checkpointinga Passing `gradient_checkpointing` to a config initialization is deprecated and will be removed in v5 Transformers. Using `model.gradient_checkpointing_enable()` instead, or if you are using the `Trainer` API, pass `gradient_checkpointing=True` in your `TrainingArguments`.z
Can't set z with value z for )5popr+ r, r- r. r/ r0 r1 r2 r3 r4 r5 r6 r7 r8 r9 _get_global_generation_defaultsitemssetattrr: r; r< r=
isinstancedict
ValueErrorlenloggerwarningr> strr torchgetattrrJ rK rL rM rN rO rP rQ rR
_name_or_pathrX _attn_implementation_internal _attn_implementation_autosetrZ getwarningswarnAttributeErrorerror)
r$ kwargsparameter_name
default_valuer> rg allowed_problem_typesr% r& errr) r) r* __init__ s
zPretrainedConfig.__init__returnc C s t | dd S )Nri )rh r$ r) r) r* rW E s z PretrainedConfig.name_or_pathc C s t || _d S N)rf ri r$ r& r) r) r* rW I s c C s | j o| j S )zY
`bool`: Whether or not return [`~utils.ModelOutput`] instead of tuples.
)r+ r. rx r) r) r* use_return_dictM s z PretrainedConfig.use_return_dictc C s
t | jS )zH
`int`: The number of labels for classification models.
)rc r< rx r) r) r* r> U s
zPretrainedConfig.num_labelsr> c C sZ t | dr| jd u st| j|kr+dd t|D | _tt| j | j | _d S d S )Nr< c S s i | ]}|d | qS )LABEL_r) )rF ir) r) r* rG _ s z/PretrainedConfig.num_labels..) hasattrr< rc rangera zipvalueskeysr= )r$ r> r) r) r* r> \ s " c C s" t | dr| jd u r dS | jS dS )Nrj eager)r~ rj rx r) r) r* _attn_implementationb s
z%PretrainedConfig._attn_implementationc C s
|| _ d S ry )rj rz r) r) r* r n s
save_directory push_to_hubc K s | | tj|rtd| d| }t|dkr(tdt | t
tj |dd |rR| dd}| d |
tjjd
}| j|fi |}| |}| jdur^t| || d tj|t}| j|dd td
| |r| j|||||dd dS dS )aS
Save a configuration object to the directory `save_directory`, so that it can be re-loaded using the
[`~PretrainedConfig.from_pretrained`] class method.
Args:
save_directory (`str` or `os.PathLike`):
Directory where the configuration JSON file will be saved (will be created if it does not exist).
push_to_hub (`bool`, *optional*, defaults to `False`):
Whether or not to push your model to the Hugging Face model hub after saving it. You can specify the
repository you want to push to with `repo_id` (will default to the name of `save_directory` in your
namespace).
kwargs (`Dict[str, Any]`, *optional*):
Additional key word arguments passed along to the [`~utils.PushToHubMixin.push_to_hub`] method.
zProvided path (z#) should be a directory, not a filer a Some non-default generation parameters are set in the model config. These should go into either a) `model.generation_config` (as opposed to `model.config`); OR b) a GenerationConfig file (https://huggingface.co/docs/transformers/generation_strategies#save-a-custom-decoding-strategy-with-your-model).This warning will become an exception in the future.
Non-default generation parameters: T)exist_okcommit_messageNrepo_id)configuse_diffzConfiguration saved in token)r r )_set_token_in_kwargsospathisfileAssertionError&_get_non_default_generation_parametersrc rm rn rf UserWarningmakedirsr\ splitsep _create_repo_get_files_timestampsr r joinr
to_json_filerd info_upload_modified_filesrl ) r$ r r rq !non_default_generation_parametersr r files_timestampsoutput_config_filer) r) r* save_pretrainedr s>
z PretrainedConfig.save_pretrainedc C s` |du r
| dd}| dd}|dur$tdt |dur"td|}|dur.|| d< dS dS )zTemporary method to deal with `token` and `use_auth_token`.
This method is to avoid apply the same changes in all model config classes that overwrite `from_pretrained`.
Need to clean up `use_auth_token` in a follow PR.
Nr use_auth_tokenzrThe `use_auth_token` argument is deprecated and will be removed in v5 of Transformers. Please use `token` instead.zV`token` and `use_auth_token` are both specified. Please set only the argument `token`.)r\ rm rn
FutureWarningrb )rq r r r) r) r* r s z%PretrainedConfig._set_token_in_kwargsmain pretrained_model_name_or_path cache_dirforce_downloadlocal_files_onlyr revisionc K s ||d< ||d< ||d< ||d< | || | j|fi |\}}| jr.| j|v r.|| j }d|v rlt| drl|d | jkrl| D ]\} }
t|
trU|
d| jkrU|
}qB|d | jkrlt
d|d d| j d | j |fi |S ) a
Instantiate a [`PretrainedConfig`] (or a derived class) from a pretrained model configuration.
Args:
pretrained_model_name_or_path (`str` or `os.PathLike`):
This can be either:
- a string, the *model id* of a pretrained model configuration hosted inside a model repo on
huggingface.co.
- a path to a *directory* containing a configuration file saved using the
[`~PretrainedConfig.save_pretrained`] method, e.g., `./my_model_directory/`.
- a path or url to a saved configuration JSON *file*, e.g., `./my_model_directory/configuration.json`.
cache_dir (`str` or `os.PathLike`, *optional*):
Path to a directory in which a downloaded pretrained model configuration should be cached if the
standard cache should not be used.
force_download (`bool`, *optional*, defaults to `False`):
Whether or not to force to (re-)download the configuration files and override the cached versions if
they exist.
resume_download:
Deprecated and ignored. All downloads are now resumed by default when possible.
Will be removed in v5 of Transformers.
proxies (`Dict[str, str]`, *optional*):
A dictionary of proxy servers to use by protocol or endpoint, e.g., `{'http': 'foo.bar:3128',
'http://hostname': 'foo.bar:4012'}.` The proxies are used on each request.
token (`str` or `bool`, *optional*):
The token to use as HTTP bearer authorization for remote files. If `True`, or not specified, will use
the token generated when running `huggingface-cli login` (stored in `~/.huggingface`).
revision (`str`, *optional*, defaults to `"main"`):
The specific model version to use. It can be a branch name, a tag name, or a commit id, since we use a
git-based system for storing models and other artifacts on huggingface.co, so `revision` can be any
identifier allowed by git.
To test a pull request you made on the Hub, you can pass `revision="refs/pr/"`.
return_unused_kwargs (`bool`, *optional*, defaults to `False`):
If `False`, then this function returns just the final configuration object.
If `True`, then this functions returns a `Tuple(config, unused_kwargs)` where *unused_kwargs* is a
dictionary consisting of the key/value pairs whose keys are not configuration attributes: i.e., the
part of `kwargs` which has not been used to update `config` and is otherwise ignored.
subfolder (`str`, *optional*, defaults to `""`):
In case the relevant files are located inside a subfolder of the model repo on huggingface.co, you can
specify the folder name here.
kwargs (`Dict[str, Any]`, *optional*):
The values in kwargs of any keys which are configuration attributes will be used to override the loaded
values. Behavior concerning key/value pairs whose keys are *not* configuration attributes is controlled
by the `return_unused_kwargs` keyword parameter.
Returns:
[`PretrainedConfig`]: The configuration object instantiated from this pretrained model.
Examples:
```python
# We can't instantiate directly the base class *PretrainedConfig* so let's show the examples on a
# derived class: BertConfig
config = BertConfig.from_pretrained(
"google-bert/bert-base-uncased"
) # Download configuration from huggingface.co and cache.
config = BertConfig.from_pretrained(
"./test/saved_model/"
) # E.g. config (or model) was saved using *save_pretrained('./test/saved_model/')*
config = BertConfig.from_pretrained("./test/saved_model/my_configuration.json")
config = BertConfig.from_pretrained("google-bert/bert-base-uncased", output_attentions=True, foo=False)
assert config.output_attentions == True
config, unused_kwargs = BertConfig.from_pretrained(
"google-bert/bert-base-uncased", output_attentions=True, foo=False, return_unused_kwargs=True
)
assert config.output_attentions == True
assert unused_kwargs == {"foo": False}
```r r r r r z You are using a model of type z to instantiate a model of type zN. This is not supported for all configurations of models and can yield errors.) r get_config_dictr r~ r r^ r` ra rl rd re from_dict) clsr r r r r r rq config_dictkvr) r) r* from_pretrained s( V
z PretrainedConfig.from_pretrainedc K s | | t|}| j|fi |\}}|du r i |fS d|v r'|d |d< d|v r>t|d }| j|fd|i|\}}||fS )a
From a `pretrained_model_name_or_path`, resolve to a dictionary of parameters, to be used for instantiating a
[`PretrainedConfig`] using `from_dict`.
Parameters:
pretrained_model_name_or_path (`str` or `os.PathLike`):
The identifier of the pre-trained checkpoint from which we want the dictionary of parameters.
Returns:
`Tuple[Dict, Dict]`: The dictionary(ies) that will be used to instantiate the configuration object.
NrX configuration_files_configuration_file)r copydeepcopy_get_config_dictget_configuration_file)r r rq original_kwargsr configuration_filer) r) r* r : s"
z PretrainedConfig.get_config_dictc K s | dd }| dd}| dd }| dd }| dd }| dd}| dd } | d d }
| d
d } | d d } | d
d}
| dd }|dd }|
du rWtd d|
d}| d urd| |d< t|}tj|}tjtj | |r~|}d}nZt
|r|d u r|n|}t |}nI|d u r| dt n|}z t
|||||||||| | |d }|d u rd |fW S t||}W n ty ty td| d| d| dw z|rt|ddd }n| |}||d< W n tjtfy td | d w |rtd | n td | d | d |v r*|s*t|d ||d < d!|v r;|s;t|d! ||d!< d"|vrIt|rId#|d"< ||fS )$Nr r Fresume_downloadproxiesr r r trust_remote_code subfolderr _from_pipeline
_from_autorX gguf_fileTzgThe argument `trust_remote_code` is to be used with Auto classes. It has no effect here and is ignored.r ) file_typefrom_auto_classusing_pipeliner )
r r r r r r
user_agentr r rX z!Can't load the configuration of 'z'. If you were trying to load it from 'https://huggingface.co/models', make sure you don't have a local directory with the same name. Otherwise, make sure 'z2' is the correct path to a directory containing a z file)return_tensorsz"It looks like the config file at 'z' is not a valid JSON file.zloading configuration file z from cache at auto_mapcustom_pipelinesr timm_wrapper)r\ rl rd re rf r r isdirr r r r r
r r OSError Exceptionr _dict_from_json_filejsonJSONDecodeErrorUnicodeDecodeErrorr r r
r )r r rq r r r r r r r r r
from_pipeliner commit_hashr r is_localresolved_config_filer r r) r) r* r ] s
z!PretrainedConfig._get_config_dictr c K sv | dd}| dd | dd d|v r d|v r |d |d< | dd|d< | di |}t|dr?d d
|j D |_d |v rjd |v rj|d }|d durU|d ng }t||krjtd
| d|d dg }| D ]/\}} t||rt||}
t|
trt| t r|
j
di | } t ||| |dkr| | qp|D ]}| |d qt
d| |r||fS |S )ao
Instantiates a [`PretrainedConfig`] from a Python dictionary of parameters.
Args:
config_dict (`Dict[str, Any]`):
Dictionary that will be used to instantiate the configuration object. Such a dictionary can be
retrieved from a pretrained checkpoint by leveraging the [`~PretrainedConfig.get_config_dict`] method.
kwargs (`Dict[str, Any]`):
Additional parameters from which to initialize the configuration object.
Returns:
[`PretrainedConfig`]: The configuration object instantiated from those parameters.
return_unused_kwargsFr Nr rX rY r2 c S rB r) rC rE r) r) r* rG rH z.PretrainedConfig.from_dict..r> r< r? r@ zX. Since those arguments are inconsistent with each other, you should remove one of them.r/ z
Model config r) )r\ r~ r2 r^ rc rb rh r` r ra r( r_ appendrd r ) r r rq r r r> r< to_remover% r& current_attrr) r) r* r sD
zPretrainedConfig.from_dict json_filec C s | |}| di |S )aQ
Instantiates a [`PretrainedConfig`] from the path to a JSON file of parameters.
Args:
json_file (`str` or `os.PathLike`):
Path to the JSON file containing the parameters.
Returns:
[`PretrainedConfig`]: The configuration object instantiated from that JSON file.
Nr) )r )r r r r) r) r* from_json_file s
z PretrainedConfig.from_json_filec C s> t |dd }| }W d n1 sw Y t|S )Nutf-8encoding)openreadr loads)r r readertextr) r) r* r s
z%PretrainedConfig._dict_from_json_filec C s t |to
| j|jkS ry )r` r __dict__)r$ otherr) r) r* __eq__ zPretrainedConfig.__eq__c C s | j j d| S )N )r( __name__to_json_stringrx r) r) r* __repr__ r zPretrainedConfig.__repr__c c s | j E d H d S ry )r rx r) r) r* __iter__ s zPretrainedConfig.__iter__c C sd | }t }| js| ni }i }| D ]Y\}}tt| |dtr2||v r2t|| ts7|| jv rQt ||t| |dd}d|v rL|d |d< |||< q||vso|dkso|dkso||| kso||v rs||
||krs|||< qt | drt| j ts| j n| j |d< |
dd}| | d|v r|d= d |v r|d = d
|v r|d
= d |v r|d = |S ) ar
Removes all attributes from the configuration that correspond to the default config attributes for
better readability, while always retaining the `config` attribute from the class. Serializes to a
Python dictionary.
Returns:
Dict[str, Any]: Dictionary of all the attributes that make up this configuration instance.
N
config_objr rZ
vocab_filequantization_config_pre_quantization_dtyperj r r ri )to_dictr r r( r^ r` rh ra r recursive_diff_dictrl r~ r r\ dict_torch_dtype_to_str) r$ r default_config_dictclass_config_dictserializable_config_dictr% r& diff_r) r) r* to_diff_dict" sL
z PretrainedConfig.to_diff_dictc C s t | j}t| jdr| jj|d< d|v r|d= d|v r |d= d|v r'|d= d|v r.|d= d|v r5|d= t|d< | D ]\}}t|t rM|
}|d= |||< q=t| drmt| j t sb| j
n| j |d< |
d d
}| | |S ) z
Serializes this instance to a Python dictionary.
Returns:
`Dict[str, Any]`: Dictionary of all the attributes that make up this configuration instance.
r r rX rj r r rZ r r N)r r r r~ r( r r r^ r` r r r ra r\ r )r$ outputr% r& r r) r) r* r f s6
zPretrainedConfig.to_dictTr c C s. |du r | }n| }tj|dddd S )a
Serializes this instance to a JSON string.
Args:
use_diff (`bool`, *optional*, defaults to `True`):
If set to `True`, only the difference between the config instance and the default `PretrainedConfig()`
is serialized to JSON string.
Returns:
`str`: String containing all the attributes that make up this configuration instance in JSON format.
TrI )indent sort_keys
)r r r dumps)r$ r r r) r) r* r s
z PretrainedConfig.to_json_stringjson_file_pathc C sF t |ddd}|| j|d W d dS 1 s w Y dS )a
Save this instance to a JSON file.
Args:
json_file_path (`str` or `os.PathLike`):
Path to the JSON file in which this configuration instance's parameters will be saved.
use_diff (`bool`, *optional*, defaults to `True`):
If set to `True`, only the difference between the config instance and the default `PretrainedConfig()`
is serialized to JSON file.
wr r r N)r writer )r$ r r writerr) r) r* r s "z PretrainedConfig.to_json_filec C s" | D ]
\}}t| || qdS )z
Updates attributes of this class with attributes from `config_dict`.
Args:
config_dict (`Dict[str, Any]`): Dictionary of attributes that should be updated for this class.
N)r^ r_ )r$ r r% r& r) r) r* update s zPretrainedConfig.update
update_strc C s t dd |dD }| D ]a\}}t| |s!td| dt| |}t|trH| dv r4d}n7| dv r=d }n.td
| d | d t|t rRt |}nt|t
r\t
|}nt|t skt d
| d| t
| || qdS )a
Updates attributes of this class with attributes from `update_str`.
The expected format is ints, floats and strings as is, and for booleans use `true` or `false`. For example:
"n_embd=10,resid_pdrop=0.2,scale_attn_weights=false,summary_type=cls_index"
The keys to change have to already exist in the config object.
Args:
update_str (`str`): String with attributes that should be updated for this class.
c s s | ]}| d V qdS )=N)r )rF xr) r) r* s z6PretrainedConfig.update_from_string..,zkey z" isn't in the original config dict)true1yyesT)false0nnoFz can't derive true or false from z (key )zIYou can only update int, float, bool or string values in the config, got z for key N)ra r r^ r~ rb rh r` boollowerrD floatrf TypeErrorr_ )r$ r dr r old_vr) r) r* update_from_string s*
z#PretrainedConfig.update_from_stringr c C s | dddur1t|d tr dd |d D |d< nt|d ts1t|d dd |d< | D ] }t|trA| | q5dS )a.
Checks whether the passed dictionary and its nested dicts have a *torch_dtype* key and if it's not None,
converts torch.dtype to a string of just the type. For example, `torch.float32` get converted into *"float32"*
string, which can then be stored in the json format.
r/ Nc S s$ i | ]\}}|t |d d qS )rA r )rf r )rF r r r) r) r* rG s $ z.rA r )rl r` ra r^ rf r r r )r$ r r& r) r) r* r s
z(PretrainedConfig.dict_torch_dtype_to_str
AutoConfigc C sD t |ts|j}ddlm m} t||s t| d|| _dS )a
Register this class with a given auto class. This should only be used for custom configurations as the ones in
the library are already mapped with `AutoConfig`.
This API is experimental and may have some slight breaking changes in the next releases.
Args:
auto_class (`str` or `type`, *optional*, defaults to `"AutoConfig"`):
The auto class to register this new configuration with.
r Nz is not a valid auto class.) r` rf r transformers.models.automodelsautor~ rb r )r
auto_class auto_moduler) r) r* register_for_auto_class s
z(PretrainedConfig.register_for_auto_classc
C s i dddddddddd d
d d d d
ddddddddddddddddd dd ddd d dd d d dS )N
max_length
min_lengthr do_sampleFearly_stopping num_beamsr num_beam_groupsdiversity_penaltyg temperatureg ?top_k2 top_p typical_prepetition_penaltylength_penaltyno_repeat_ngram_size encoder_no_repeat_ngram_size
bad_words_idsnum_return_sequences)
output_scoresreturn_dict_in_generateforced_bos_token_idforced_eos_token_idremove_invalid_values exponential_decay_length_penaltysuppress_tokensbegin_suppress_tokensr) r) r) r) r* r] sX
z0PretrainedConfig._get_global_generation_defaultsc C s i }d}z| }W n ty$ | jdd}|| ur | }nd}Y nw |du r+| nt| |}| D ]<\}}t||rrd }} t||}
|
du rMq6|durY|
t||k}n|
|k} |du ph|du oh| du } | rrt||||< q6|S )z]
Gets the non-default generation parameters on the PretrainedConfig instance
NT)decoderF)r( rb get_text_configrh r] r^ r~ ) r$ r decoder_attribute_namedefault_configdecoder_configself_decoder_configrr default_global_valueis_default_in_configis_default_generation_valueparameter_valueis_non_defaultr) r) r* r - s8
z7PretrainedConfig._get_non_default_generation_parametersc C s d}d}|r |}n|| }g }|D ]}t | |r't| |d}|dur'||g7 }qt|dkr6td| dt|dkrEt| |d }|S | }|S )as
Returns the config that is meant to be used with text IO. On most models, it is the original config instance
itself. On specific composite models, it is under a set of valid names.
Args:
decoder (`Optional[bool]`, *optional*, defaults to `False`):
If set to `True`, then only search for decoder config names.
)r: generator text_config) text_encoderNr z r r r PathLiker staticmethodr classmethodr r r r r r r r r r r r r r r r r r] r r;
__classcell__r) r) r' r* r / s
q
< p " h < D0
& ,r r rw c C s i }| D ] }| dr |dr |dkr |dd}|||< qt| }t}tt }|D ]}t||kr>|| }q0 |S |S )z
Get the configuration file to use for this version of transformers.
Args:
configuration_files (`List[str]`): The list of available configuration files.
Returns:
`str`: The configuration file to use.
zconfig.z.jsonz config.json)
startswithendswith removeprefix removesuffixsortedr r
r parser )r configuration_files_map file_namer available_versionsr rZ r) r) r* r | s
r c C s i }|dur | ni }| D ]8\}}t|t|d}t|tr<||v r s< 4
U