sgJdZddlZddlZddlZddlZddlmZddlmZddl m Z m Z ddl m Z mZmZmZmZmZmZmZmZdd lmZmZmZerdd lmZej8eZd Zd Z d Z!dZ"dZ#dZ$GddZ%Gdde%Z&ddZ'ddZ(dZ)dZ*dZ+GddeZ,y)z-Factory function to build auto-model classes.N) OrderedDict)PretrainedConfig)get_class_from_dynamic_moduleresolve_trust_remote_code) CONFIG_NAME cached_file copy_funcextract_commit_hashfind_adapter_config_fileis_peft_availableis_torch_availableloggingrequires_backends) AutoConfigmodel_type_to_module_name!replace_list_option_in_docstrings)GenerationMixinaJ This is a generic model class that will be instantiated as one of the model classes of the library when created with the [`~BaseAutoModelClass.from_pretrained`] class method or the [`~BaseAutoModelClass.from_config`] class method. This class cannot be instantiated directly using `__init__()` (throws an error). ax Instantiates one of the model classes of the library from a configuration. Note: Loading a model from its configuration file does **not** load the model weights. It only affects the model's configuration. Use [`~BaseAutoModelClass.from_pretrained`] to load the model weights. Args: config ([`PretrainedConfig`]): The model class to instantiate is selected based on the configuration class: List options attn_implementation (`str`, *optional*): The attention implementation to use in the model (if relevant). Can be any of `"eager"` (manual implementation of the attention), `"sdpa"` (using [`F.scaled_dot_product_attention`](https://pytorch.org/docs/master/generated/torch.nn.functional.scaled_dot_product_attention.html)), or `"flash_attention_2"` (using [Dao-AILab/flash-attention](https://github.com/Dao-AILab/flash-attention)). By default, if available, SDPA will be used for torch>=2.1.1. The default is otherwise the manual `"eager"` implementation. Examples: ```python >>> from transformers import AutoConfig, BaseAutoModelClass >>> # Download configuration from huggingface.co and cache. >>> config = AutoConfig.from_pretrained("checkpoint_placeholder") >>> model = BaseAutoModelClass.from_config(config) ``` ac Instantiate one of the model classes of the library from a pretrained model. The model class to instantiate is selected based on the `model_type` property of the config object (either passed as an argument or loaded from `pretrained_model_name_or_path` if possible), or when it's missing, by falling back to using pattern matching on `pretrained_model_name_or_path`: List options The model is set in evaluation mode by default using `model.eval()` (so for instance, dropout modules are deactivated). To train the model, you should first set it back in training mode with `model.train()` Args: pretrained_model_name_or_path (`str` or `os.PathLike`): Can be either: - A string, the *model id* of a pretrained model hosted inside a model repo on huggingface.co. - A path to a *directory* containing model weights saved using [`~PreTrainedModel.save_pretrained`], e.g., `./my_model_directory/`. - A path or url to a *tensorflow index checkpoint file* (e.g, `./tf_model/model.ckpt.index`). In this case, `from_tf` should be set to `True` and a configuration object should be provided as `config` argument. This loading path is slower than converting the TensorFlow checkpoint in a PyTorch model using the provided conversion scripts and loading the PyTorch model afterwards. model_args (additional positional arguments, *optional*): Will be passed along to the underlying model `__init__()` method. config ([`PretrainedConfig`], *optional*): Configuration for the model to use instead of an automatically loaded configuration. Configuration can be automatically loaded when: - The model is a model provided by the library (loaded with the *model id* string of a pretrained model). - The model was saved using [`~PreTrainedModel.save_pretrained`] and is reloaded by supplying the save directory. - The model is loaded by supplying a local directory as `pretrained_model_name_or_path` and a configuration JSON file named *config.json* is found in the directory. state_dict (*Dict[str, torch.Tensor]*, *optional*): A state dictionary to use instead of a state dictionary loaded from saved weights file. This option can be used if you want to create a model from a pretrained configuration but load your own weights. In this case though, you should check if using [`~PreTrainedModel.save_pretrained`] and [`~PreTrainedModel.from_pretrained`] is not a simpler option. 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. from_tf (`bool`, *optional*, defaults to `False`): Load the model weights from a TensorFlow checkpoint save file (see docstring of `pretrained_model_name_or_path` argument). force_download (`bool`, *optional*, defaults to `False`): Whether or not to force the (re-)download of the model weights and configuration files, overriding 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. output_loading_info(`bool`, *optional*, defaults to `False`): Whether ot not to also return a dictionary containing missing keys, unexpected keys and error messages. local_files_only(`bool`, *optional*, defaults to `False`): Whether or not to only look at local files (e.g., not try downloading the model). 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. trust_remote_code (`bool`, *optional*, defaults to `False`): Whether or not to allow for custom models defined on the Hub in their own modeling files. This option should only be set to `True` for repositories you trust and in which you have read the code, as it will execute code present on the Hub on your local machine. code_revision (`str`, *optional*, defaults to `"main"`): The specific revision to use for the code on the Hub, if the code leaves in a different repository than the rest of the model. 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. kwargs (additional keyword arguments, *optional*): Can be used to update the configuration object (after it being loaded) and initiate the model (e.g., `output_attentions=True`). Behaves differently depending on whether a `config` is provided or automatically loaded: - If a configuration is provided with `config`, `**kwargs` will be directly passed to the underlying model's `__init__` method (we assume all relevant updates to the configuration have already been done) - If a configuration is not provided, `kwargs` will be first passed to the configuration class initialization function ([`~PretrainedConfig.from_pretrained`]). Each key of `kwargs` that corresponds to a configuration attribute will be used to override said attribute with the supplied `kwargs` value. Remaining keys that do not correspond to any configuration attribute will be passed to the underlying model's `__init__` function. Examples: ```python >>> from transformers import AutoConfig, BaseAutoModelClass >>> # Download model and configuration from huggingface.co and cache. >>> model = BaseAutoModelClass.from_pretrained("checkpoint_placeholder") >>> # Update configuration during loading >>> model = BaseAutoModelClass.from_pretrained("checkpoint_placeholder", output_attentions=True) >>> model.config.output_attentions True >>> # Loading from a TF checkpoint file instead of a PyTorch model (slower) >>> config = AutoConfig.from_pretrained("./tf_model/shortcut_placeholder_tf_model_config.json") >>> model = BaseAutoModelClass.from_pretrained( ... "./tf_model/shortcut_placeholder_tf_checkpoint.ckpt.index", from_tf=True, config=config ... ) ``` a Instantiate one of the model classes of the library from a pretrained model. The model class to instantiate is selected based on the `model_type` property of the config object (either passed as an argument or loaded from `pretrained_model_name_or_path` if possible), or when it's missing, by falling back to using pattern matching on `pretrained_model_name_or_path`: List options Args: pretrained_model_name_or_path (`str` or `os.PathLike`): Can be either: - A string, the *model id* of a pretrained model hosted inside a model repo on huggingface.co. - A path to a *directory* containing model weights saved using [`~PreTrainedModel.save_pretrained`], e.g., `./my_model_directory/`. - A path or url to a *PyTorch state_dict save file* (e.g, `./pt_model/pytorch_model.bin`). In this case, `from_pt` should be set to `True` and a configuration object should be provided as `config` argument. This loading path is slower than converting the PyTorch model in a TensorFlow model using the provided conversion scripts and loading the TensorFlow model afterwards. model_args (additional positional arguments, *optional*): Will be passed along to the underlying model `__init__()` method. config ([`PretrainedConfig`], *optional*): Configuration for the model to use instead of an automatically loaded configuration. Configuration can be automatically loaded when: - The model is a model provided by the library (loaded with the *model id* string of a pretrained model). - The model was saved using [`~PreTrainedModel.save_pretrained`] and is reloaded by supplying the save directory. - The model is loaded by supplying a local directory as `pretrained_model_name_or_path` and a configuration JSON file named *config.json* is found in the directory. 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. from_pt (`bool`, *optional*, defaults to `False`): Load the model weights from a PyTorch checkpoint save file (see docstring of `pretrained_model_name_or_path` argument). force_download (`bool`, *optional*, defaults to `False`): Whether or not to force the (re-)download of the model weights and configuration files, overriding 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. output_loading_info(`bool`, *optional*, defaults to `False`): Whether ot not to also return a dictionary containing missing keys, unexpected keys and error messages. local_files_only(`bool`, *optional*, defaults to `False`): Whether or not to only look at local files (e.g., not try downloading the model). 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. trust_remote_code (`bool`, *optional*, defaults to `False`): Whether or not to allow for custom models defined on the Hub in their own modeling files. This option should only be set to `True` for repositories you trust and in which you have read the code, as it will execute code present on the Hub on your local machine. code_revision (`str`, *optional*, defaults to `"main"`): The specific revision to use for the code on the Hub, if the code leaves in a different repository than the rest of the model. 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. kwargs (additional keyword arguments, *optional*): Can be used to update the configuration object (after it being loaded) and initiate the model (e.g., `output_attentions=True`). Behaves differently depending on whether a `config` is provided or automatically loaded: - If a configuration is provided with `config`, `**kwargs` will be directly passed to the underlying model's `__init__` method (we assume all relevant updates to the configuration have already been done) - If a configuration is not provided, `kwargs` will be first passed to the configuration class initialization function ([`~PretrainedConfig.from_pretrained`]). Each key of `kwargs` that corresponds to a configuration attribute will be used to override said attribute with the supplied `kwargs` value. Remaining keys that do not correspond to any configuration attribute will be passed to the underlying model's `__init__` function. Examples: ```python >>> from transformers import AutoConfig, BaseAutoModelClass >>> # Download model and configuration from huggingface.co and cache. >>> model = BaseAutoModelClass.from_pretrained("checkpoint_placeholder") >>> # Update configuration during loading >>> model = BaseAutoModelClass.from_pretrained("checkpoint_placeholder", output_attentions=True) >>> model.config.output_attentions True >>> # Loading from a PyTorch checkpoint file instead of a TensorFlow model (slower) >>> config = AutoConfig.from_pretrained("./pt_model/shortcut_placeholder_pt_model_config.json") >>> model = BaseAutoModelClass.from_pretrained( ... "./pt_model/shortcut_placeholder_pytorch_model.bin", from_pt=True, config=config ... ) ``` c|t|}t|ttfs|S|Dcic]}|j|}}t |dg}|D]/}||vr||cSd||vr |d|cSd||vs'|d|cS|dScc}w)N architecturesTFFlaxr)type isinstancelisttuple__name__getattr)config model_mappingsupported_modelsmodel name_to_modelrarchs X/var/www/html/venv/lib/python3.12/site-packages/transformers/models/auto/auto_factory.py_get_model_classr's$T&\2 &u 68HIuU^^U*IMIFOR8M0 =  & &$[M ) 2dV- -D6]m + 4v/ / 0 A JsBcHeZdZdZdZedZedZeddZy)_BaseAutoModelClassNct|jjd|jjd|jjd)Nz+ is designed to be instantiated using the `z5.from_pretrained(pretrained_model_name_or_path)` or `z.from_config(config)` methods.)EnvironmentError __class__r)selfargskwargss r&__init__z_BaseAutoModelClass.__init__sR~~&&'(..1123''((F H  c p|jdd}t|dxr|j|jv}t ||j j v}t||j||}|r|r|j|j}d|vr|jd\}}n |j}t||fi|}|j|j|d|jdd} t|}|j|fi|St ||j j vr)t!||j }|j|fi|St#d|jd|jd d j%d |j j Dd ) Ntrust_remote_codeauto_mapz--Texist_ok code_revision!Unrecognized configuration class  for this kind of AutoModel: . Model type should be one of , c34K|]}|jywNr.0cs r& z2_BaseAutoModelClass.from_config..4cAQZZ4c.)pophasattrrr4r_model_mappingkeysr _name_or_pathsplit name_or_pathrregisterr,$add_generation_mixin_to_remote_model _from_configr' ValueErrorjoin) clsr r/r3has_remote_codehas_local_code class_refrepo_id model_class_s r& from_configz_BaseAutoModelClass.from_configs"JJ':DA!&*5Y#,,&//:Yf););)@)@)BB5 v33^_  0 5Iy %.__T%:" --7 7UfUK LL));L F ?D1A>{KK+;++F=f= = &\S//446 6*633E3EFK+;++F=f= =/0@0@/AA^_b_k_k^lm++/994cI[I[I`I`Ib4c+c*dde g  r1c |jdd}|jdd}d|d<gd}|Dcic]}||vs||j|}}|jdd} |jdd} |jdd} |jd d} |jd d} | )tjd t| t d | } | | |d <| ?t |t s"t|tfd d d d|}t|| } n t|dd} trQ| i} | | | d <t|fd| i| }|6t|dd5}tj|}|| d<|d}dddt |t st!j"|}|j%dddk(r|jd}|j%dd|jd}t'j(|fd|| | d||\}}|j%dddk(rd|d<|j%dd|d|d<t+|dxr|j,|j.v}t1||j2j5v}t7||||}| |d<|r|r|j.|j,}t9||fd| i||}|jdd}|j;|j<|dt?|}|j(|g|d|i||St1||j2j5vr0tA||j2}|j(|g|d|i||St d|j<d|j,ddjCd|j2j5Ddcc}w#1swY^xYw) Nr r3T _from_auto) cache_dirforce_downloadlocal_files_onlyproxiesresume_downloadrevision subfolderuse_auth_tokentokenr7 _commit_hashadapter_kwargsrdrczrThe `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`.F) _raise_exceptions_for_gated_repo%_raise_exceptions_for_missing_entries'_raise_exceptions_for_connection_errorsrzutf-8)encoding_adapter_model_pathbase_model_name_or_path torch_dtypeautoquantization_config)return_unused_kwargsr3r7rer4r5r8r9r:r;c34K|]}|jywr=r>r?s r&rBz6_BaseAutoModelClass.from_pretrained..9rCrDrE)"rFwarningswarn FutureWarningrPrrr rr rr r openjsonloadcopydeepcopygetrfrom_pretrainedrGrr4rrHrIrrrMr,rNr'rQ)rRpretrained_model_name_or_path model_argsr/r r3hub_kwargs_namesname hub_kwargsr7 commit_hashrfrdrcresolved_config_filemaybe_adapter_pathfadapter_config kwargs_origrXrSrTrUrWs r&r|z#_BaseAutoModelClass.from_pretrainedsHd+"JJ':DA#|  :J\TU[^dFJJt,,\ \ ?D9 jj6 $4d;w-#(8$?  % MME   l#E  "'Jw   f&67'21(6;:?{KK.;..-0:CIMW[a &\S//446 6*633E3EFK.;..-0:CIMW[a /0@0@/AA^_b_k_k^lm++/994cI[I[I`I`Ib4c+c*dde g  Q]^^^s O O  OOct|dr=t|jt|k7rtd|jd|d|jj |||y)a Register a new model for this class. Args: config_class ([`PretrainedConfig`]): The configuration corresponding to the model to register. model_class ([`PreTrainedModel`]): The model to register. config_classzThe model class you are passing has a `config_class` attribute that is not consistent with the config class you passed (model has z and you passed z!. Fix one of those so they match!r5N)rGstrrrPrHrM)rRrrWr6s r&rMz_BaseAutoModelClass.register<sv ; /C 8P8P4QUXYeUf4f66A6N6N5OO_`l_mn..  ##L+#Qr1F) r __module__ __qualname__rHr0 classmethodrYr|rMr1r&r)r)sMN   8z z xRRr1r)c@eZdZdZefdZefdZxZS)_BaseAutoBackboneClassNct|ddgddlm}|jd|}|j dd t d|j dd r t d |jd |j }|jd |j}|jd |j}|jd|j} |||||| }t |,|fi|S)Nvisiontimmr)TimmBackboneConfigr out_featuresz0Cannot specify `out_features` for timm backbonesoutput_loading_infoFz@Cannot specify `output_loading_info=True` when loading from timm num_channels features_onlyuse_pretrained_backbone out_indices)backbonerrrr) rmodels.timm_backbonerrFr{rPrrrrsuperrY) rRr}r~r/rr rrrrr,s r&#_load_timm_backbone_from_pretrainedz:_BaseAutoBackboneClass._load_timm_backbone_from_pretrainedTs#&12>H&8&:; ::nd + 7OP P ::+U 3_` `zz.&2E2EF  ?F4H4HI "(**-FHfHf"gjj0B0BC #2%'$;#  w"64V44r1c~|jdd}|r|j|g|i|St| |g|i|S)Nuse_timm_backboneF)rFrrr|)rRr}r~r/rr,s r&r|z&_BaseAutoBackboneClass.from_pretrainednsT"JJ':EB :3::;Xp[epiop pw&'D\z\U[\\r1)rrrrHrrr| __classcell__)r,s@r&rrPs0N552]]r1rcnt|dkDr|jdd|dS|jddS)Nrz(one of the model classes of the library z0one of the model classes of the library (with a z head) z-one of the base model classes of the library )lenreplace) docstringhead_docs r&insert_head_docrwsK 8}q  6>xj P     24c r1c^|j}|j}tt|}|j d||_t tj}tt|}|j d|}|j d|}||_t|jd|}t||_|jdrt}n|jdrt}nt}t tj } t||}|j d|}|j d|}|j#dd j#d d } |j d | }|| _t|j| } t| |_|S) N)rBaseAutoModelClasscheckpoint_placeholderF)use_model_typesrr/-rshortcut_placeholder)rHrrCLASS_DOCSTRINGr__doc__r r)rYFROM_CONFIG_DOCSTRINGrr startswithFROM_PRETRAINED_TF_DOCSTRINGFROM_PRETRAINED_FLAX_DOCSTRINGFROM_PRETRAINED_TORCH_DOCSTRINGr|rK) rRcheckpoint_for_examplerr!rclass_docstringrYfrom_config_docstringfrom_pretrained_docstringr|shortcuts r&auto_class_updaters&&M <ECK/;;.sGQ-fa8Gs transformerszCould not find z neither in z nor in !z in )rrrGr importlib import_modulerrP)rattrtransformers_modules` r&rrs |$G$GGGvtvt$$$11.A $$ i+,?F F?4&5H4IKLL itfLQdPeefgh h is  A::Bc$dt|jvr|Sdt|jvr|Sdtt|dv}dtt|dv}|s|r+t |j |t fi|j}|S|S)a Adds `GenerationMixin` to the inheritance of `model_class`, if `model_class` is a PyTorch model. This function is used for backwards compatibility purposes: in v4.45, we've started a deprecation cycle to make `PreTrainedModel` stop inheriting from `GenerationMixin`. Without this function, older models dynamically loaded from the Hub may not have the `generate` method after we remove the inheritance. ztorch.nn.modules.module.Modulergenerateprepare_inputs_for_generation)r__mro__ __bases__rrrr__dict__)rWhas_custom_generatehas_custom_prepare_inputs!model_class_with_generation_mixins r&rNrNs(s;3F3F/GGC 5 566,3w{J7W3XX 1W[Rq=r9s s7,0  ;"@BZ[EYEYBZ- )10 r1cZeZdZdZdZdZdZdZdZdZ dZ d Z d Z d Z d Zdd Zy)_LazyAutoMappinga " A mapping config to object (model or tokenizer for instance) that will load keys and values when it is accessed. Args: - config_mapping: The map model type to config class - model_mapping: The map model type to model (or tokenizer) class c||_|jDcic]\}}|| c}}|_||_||j_i|_i|_ycc}}wr=)_config_mappingitems_reverse_config_mappingrH_extra_content_modules)r-config_mappingr!kvs r&r0z_LazyAutoMapping.__init__sY-9G9M9M9O'PA1'P$+-1*  (Qs Act|jjj|jj}t |t |j zSr=)setrrI intersectionrHrr)r- common_keyss r&__len__z_LazyAutoMapping.__len__sP$..3356CCDDWDWD\D\D^_ ;#d&9&9":::r1c||jvr|j|S|j|j}||jvr!|j|}|j ||S|j j Dcgc]\}}||jk(s|}}}|D]3}||jvs|j|}|j ||cSt|cc}}wr=)rrrrH_load_attr_from_modulerrKeyError)r-key model_type model_namerr model_typesmtypes r& __getitem__z_LazyAutoMapping.__getitem__s $%% %&&s+ +11#,,? ,, ,,,Z8J..z:F F&*%9%9%?%?%AWTQQ#,,EVqW W  FE+++!007 225*EE Fsm Xs C%C%ct|}||jvr&tjd|d|j|<t |j||S)NrEztransformers.models)rrrrr)r-rr module_names r&rz'_LazyAutoMapping._load_attr_from_modulesQ/ ; dmm +)2)@)@1[MARTi)jDMM+ &' k(BDIIr1c |jjDcgc]3\}}||jjvr|j ||5}}}|t |j jzScc}}wr=)rrrHrIrrr)r-rr mapping_keyss r&rIz_LazyAutoMapping.keys s~"11779 Td))..00  ' 'T 2  d4#6#6#;#;#=>>>  8A?cH |j|S#t$r|cYSwxYwr=)rr)r-rdefaults r&r{z_LazyAutoMapping.gets, ##C( ( N s  !!c4t|jSr=)boolrIr-s r&__bool__z_LazyAutoMapping.__bool__DIIK  r1c |jjDcgc]3\}}||jjvr|j ||5}}}|t |j jzScc}}wr=)rHrrrIrrrr)r-rrmapping_valuess r&rz_LazyAutoMapping.valuess~"00668 Td**//11  ' 'T 2  T%8%8%?%?%A BBB  rc V|jjDcgc]\}||jjvr>|j||j||j||j|f^}}|t |j j zScc}wr=)rHrIrrrrr)r-r mapping_itemss r&rz_LazyAutoMapping.items$s **//1  d**//11 ++C1E1Ec1JK++C1D1DS1IJ   tD$7$7$=$=$?@@@ sA!B&c4t|jSr=)iterrIrs r&__iter__z_LazyAutoMapping.__iter__/rr1c||jvryt|dr|j|jvry|j|j}||jvS)NTrF)rrGrrrH)r-itemrs r& __contains__z_LazyAutoMapping.__contains__2sV 4&& &tZ(DMMA]A],]11$--@ T0000r1ct|dr^|j|jvrF|j|j}||jj vr|st d|d||j |<y)z7 Register a new model in this mapping. r'z*' is already used by a Transformers model.N)rGrrrHrIrPr)r-rvaluer6rs r&rMz_LazyAutoMapping.register:sq 3 # 8T8T(T55cllCJT005577 1SE)S!TUU#(C r1Nr)rrrrr0rrrrIr{rrrrrrMrr1r&rrsF; J ? !C A!1 )r1r))zgoogle-bert/bert-base-casedr )-rryrrwrs collectionsrconfiguration_utilsrdynamic_module_utilsrrutilsrr r r r r rrrconfiguration_autorrr generationr get_loggerrloggerrrrrrr'r)rrrrrrNrrr1r&rs4  #3\   ih-   H %4j#Xa Fa"H(vRvRr$]0$]N FM(8c){c)r1