RЦiSSKrSSKJr SSKJrJrJrJrJr SSK J r SSK J r SSK Jr SSKJr SS KJrJr SS KJr SS KJrJrJr /S QrS \S\\\\44SjrSS \S\S\4SjjrSS \S\S\\\\\\4\4S\\\\4S\\\\4S\\\\4S\\S\\S\\S\S\ R>4Sjjr g)N)Path)AnyDictOptionalTupleUnion)urlsplit)nn)set_layer_config)load_checkpoint)load_model_config_from_hfload_model_config_from_path) PretrainedCfg)is_modelmodel_entrypointsplit_model_name_tag)parse_model_namesafe_model_name create_model model_namereturncURS5(aURSS5n[U5nURS;deURS:XaURUR4$URS:XaURUR4$[ RR UR5SnSU4$)z;Parse source and name from potentially prefixed model name.hf_hubhf-hub)r local-dirrN) startswithreplacer schemepathossplit)rparseds S/var/www/html/ai-image-ml/venv/lib/python3.13/site-packages/timm/models/_factory.pyrrsX&&''(; j !F ==7 77 7 }} }}fkk)) + %}}fkk))WW]]6;;/3 Z remove_sourcec^S[S[4SjnU(a[U5SnU"U5$)z)Return a filename / path safe model name.namercPSRSU55RS5$)Nrc3R# UHoR5(aUOSv M g7f)_N)isalnum).0cs r& 5safe_model_name..make_safe..&s?$QIIKKqS0$s%'r-)joinrstrip)r*s r& make_safe"safe_model_name..make_safe%s"ww?$??FFsKKr'r)strr)rr(r5s r&rr#s5LLL%j1"5 Z  r' pretrainedpretrained_cfgpretrained_cfg_overlaycheckpoint_path cache_dir scriptable exportableno_jitkwargsc U R5V V s0sH upU cM X_M n n n [U5upU (asU(aS5eU S:Xa[U US9up nOU S:Xa[U 5up nO SU 35eU(a*UR5HupU R X5 M O[ U 5upU(a U(dUn[ U5(d[SU-5e[U5n[XgUS9 U"S UUUUS .U D6nSSS5 U(a [WU5 W$s sn n f!,(df  N)=f) aCreate a model. Lookup model's entrypoint function and pass relevant args to create a new model. Tip: **kwargs will be passed through entrypoint fn to ``timm.models.build_model_with_cfg()`` and then the model class __init__(). kwargs values set to None are pruned before passing. Args: model_name: Name of model to instantiate. pretrained: If set to `True`, load pretrained ImageNet-1k weights. pretrained_cfg: Pass in an external pretrained_cfg for model. pretrained_cfg_overlay: Replace key-values in base pretrained_cfg with these. checkpoint_path: Path of checkpoint to load _after_ the model is initialized. cache_dir: Override model cache dir for Hugging Face Hub and Torch checkpoints. scriptable: Set layer config so that model is jit scriptable (not working for all models yet). exportable: Set layer config so that model is traceable / ONNX exportable (not fully impl/obeyed yet). no_jit: Set layer config so that model doesn't utilize jit scripted layers (so far activations only). Keyword Args: drop_rate (float): Classifier dropout rate for training. drop_path_rate (float): Stochastic depth drop rate for training. global_pool (str): Classifier global pooling type. Example: ```py >>> from timm import create_model >>> # Create a MobileNetV3-Large model with no pretrained weights. >>> model = create_model('mobilenetv3_large_100') >>> # Create a MobileNetV3-Large model with pretrained weights. >>> model = create_model('mobilenetv3_large_100', pretrained=True) >>> model.num_classes 1000 >>> # Create a MobileNetV3-Large model with pretrained weights and a new head with 10 classes. >>> model = create_model('mobilenetv3_large_100', pretrained=True, num_classes=10) >>> model.num_classes 10 >>> # Create a Dinov2 small model with pretrained weights and save weights in a custom directory. >>> model = create_model('vit_small_patch14_dinov2.lvd142m', pretrained=True, cache_dir="/data/my-models") >>> # Data will be stored at `/data/my-models/models--timm--vit_small_patch14_dinov2.lvd142m/` ``` NzKpretrained_cfg should not be set when sourcing model from Hugging Face Hub.r)r<rzUnknown model_source zUnknown model (%s))r=r>r?)r8r9r:r<) itemsrrr setdefaultrr RuntimeErrorrr r )rr8r9r:r;r<r=r>r?r@kv model_sourcemodel_id model_argspretrained_tag create_fnmodels r&rr,sW| &||~ ?~tqdad~F ?-j9L!p#pp! 8 #6O#6 2N [ (5P6 2N  A1,@ @5 "((*!!!'+&:(%C" .+N J  /*<== ,I Zv V !)#9     W/ LW@> W Vs D,D,D22 E)T)FNNNNNNN)!r#pathlibrtypingrrrrr urllib.parser torchr timm.layersr _helpersr _hubrr _pretrainedr _registryrrr__all__r7rboolrModulerrBr'r&rZsS 44!(%H&GG B  x}c/A)B "!!D!C!!NR;?6:04%)%)!%iii!sDcNM'I!JKi!)c3h 8 i "%T "23 i E#t),- iTNiTNiiiYYir'