►
KerasHub: Pretrained Models /
KerasHub pretrained models
KerasHub pretrained models
Below, we list all presets available in the KerasHub library. For more detailed usage, browse the docstring for a particular class. For an in depth introduction to our API, see the getting started guide.
The following preset names correspond to a config and weights for a pretrained
model. Any task, preprocessor, backbone, or tokenizer from_preset() can be used
to create a model from the saved preset.
backbone = keras_hub.models.Backbone.from_preset("bert_base_en")
tokenizer = keras_hub.models.Tokenizer.from_preset("bert_base_en")
classifier = keras_hub.models.TextClassifier.from_preset("bert_base_en", num_classes=2)
preprocessor = keras_hub.models.TextClassifierPreprocessor.from_preset("bert_base_en")
| Preset | Model API | Parameters | Description |
|---|---|---|---|
| albert_base_en_uncased | Albert | 11.68M | 12-layer ALBERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus. |
| albert_large_en_uncased | Albert | 17.68M | 24-layer ALBERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus. |
| albert_extra_large_en_uncased | Albert | 58.72M | 24-layer ALBERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus. |
| albert_extra_extra_large_en_uncased | Albert | 222.60M | 12-layer ALBERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus. |
| bart_base_en | Bart | 139.42M | 6-layer BART model where case is maintained. Trained on BookCorpus, English Wikipedia and CommonCrawl. |
| bart_large_en | Bart | 406.29M | 12-layer BART model where case is maintained. Trained on BookCorpus, English Wikipedia and CommonCrawl. |
| bart_large_en_cnn | Bart | 406.29M | The bart_large_en backbone model fine-tuned on the CNN+DM summarization dataset. |
| bert_tiny_en_uncased | Bert | 4.39M | 2-layer BERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus. |
| bert_tiny_en_uncased_sst2 | Bert | 4.39M | The bert_tiny_en_uncased backbone model fine-tuned on the SST-2 sentiment analysis dataset. |
| bert_small_en_uncased | Bert | 28.76M | 4-layer BERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus. |
| bert_medium_en_uncased | Bert | 41.37M | 8-layer BERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus. |
| bert_base_zh | Bert | 102.27M | 12-layer BERT model. Trained on Chinese Wikipedia. |
| bert_base_en | Bert | 108.31M | 12-layer BERT model where case is maintained. Trained on English Wikipedia + BooksCorpus. |
| bert_base_en_uncased | Bert | 109.48M | 12-layer BERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus. |
| bert_base_multi | Bert | 177.85M | 12-layer BERT model where case is maintained. Trained on trained on Wikipedias of 104 languages |
| bert_large_en | Bert | 333.58M | 24-layer BERT model where case is maintained. Trained on English Wikipedia + BooksCorpus. |
| bert_large_en_uncased | Bert | 335.14M | 24-layer BERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus. |
| bloom_560m_multi | Bloom | 559.21M | 24-layer Bloom model with hidden dimension of 1024. trained on 45 natural languages and 12 programming languages. |
| bloomz_560m_multi | Bloom | 559.21M | 24-layer Bloom model with hidden dimension of 1024. finetuned on crosslingual task mixture (xP3) dataset. |
| bloom_1.1b_multi | Bloom | 1.07B | 24-layer Bloom model with hidden dimension of 1536. trained on 45 natural languages and 12 programming languages. |
| bloomz_1.1b_multi | Bloom | 1.07B | 24-layer Bloom model with hidden dimension of 1536. finetuned on crosslingual task mixture (xP3) dataset. |
| bloom_1.7b_multi | Bloom | 1.72B | 24-layer Bloom model with hidden dimension of 2048. trained on 45 natural languages and 12 programming languages. |
| bloomz_1.7b_multi | Bloom | 1.72B | 24-layer Bloom model with hidden dimension of 2048. finetuned on crosslingual task mixture (xP3) dataset. |
| bloom_3b_multi | Bloom | 3.00B | 30-layer Bloom model with hidden dimension of 2560. trained on 45 natural languages and 12 programming languages. |
| bloomz_3b_multi | Bloom | 3.00B | 30-layer Bloom model with hidden dimension of 2560. finetuned on crosslingual task mixture (xP3) dataset. |
| clip_vit_base_patch16 | - | 149.62M | 150 million parameter, 12-layer for vision and 12-layer for text, patch size of 16, CLIP model. |
| clip_vit_base_patch32 | - | 151.28M | 151 million parameter, 12-layer for vision and 12-layer for text, patch size of 32, CLIP model. |
| clip_vit_b_32_laion2b_s34b_b79k | - | 151.28M | 151 million parameter, 12-layer for vision and 12-layer for text, patch size of 32, Open CLIP model. |
| clip_vit_large_patch14 | - | 427.62M | 428 million parameter, 24-layer for vision and 12-layer for text, patch size of 14, CLIP model. |
| clip_vit_large_patch14_336 | - | 427.94M | 428 million parameter, 24-layer for vision and 12-layer for text, patch size of 14, image size of 336, CLIP model. |
| clip_vit_h_14_laion2b_s32b_b79k | - | 986.11M | 986 million parameter, 32-layer for vision and 24-layer for text, patch size of 14, Open CLIP model. |
| clip_vit_g_14_laion2b_s12b_b42k | - | 1.37B | 1.4 billion parameter, 40-layer for vision and 24-layer for text, patch size of 14, Open CLIP model. |
| clip_vit_bigg_14_laion2b_39b_b160k | - | 2.54B | 2.5 billion parameter, 48-layer for vision and 32-layer for text, patch size of 14, Open CLIP model. |
| deberta_v3_extra_small_en | DebertaV3 | 70.68M | 12-layer DeBERTaV3 model where case is maintained. Trained on English Wikipedia, BookCorpus and OpenWebText. |
| deberta_v3_small_en | DebertaV3 | 141.30M | 6-layer DeBERTaV3 model where case is maintained. Trained on English Wikipedia, BookCorpus and OpenWebText. |
| deberta_v3_base_en | DebertaV3 | 183.83M | 12-layer DeBERTaV3 model where case is maintained. Trained on English Wikipedia, BookCorpus and OpenWebText. |
| deberta_v3_base_multi | DebertaV3 | 278.22M | 12-layer DeBERTaV3 model where case is maintained. Trained on the 2.5TB multilingual CC100 dataset. |
| deberta_v3_large_en | DebertaV3 | 434.01M | 24-layer DeBERTaV3 model where case is maintained. Trained on English Wikipedia, BookCorpus and OpenWebText. |
| deeplab_v3_plus_resnet50_pascalvoc | DeepLabV3 | 39.19M | DeepLabV3+ model with ResNet50 as image encoder and trained on augmented Pascal VOC dataset by Semantic Boundaries Dataset(SBD)which is having categorical accuracy of 90.01 and 0.63 Mean IoU. |
| densenet_121_imagenet | DenseNet | 7.04M | 121-layer DenseNet model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| densenet_169_imagenet | DenseNet | 12.64M | 169-layer DenseNet model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| densenet_201_imagenet | DenseNet | 18.32M | 201-layer DenseNet model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| distil_bert_base_en | DistilBert | 65.19M | 6-layer DistilBERT model where case is maintained. Trained on English Wikipedia + BooksCorpus using BERT as the teacher model. |
| distil_bert_base_en_uncased | DistilBert | 66.36M | 6-layer DistilBERT model where all input is lowercased. Trained on English Wikipedia + BooksCorpus using BERT as the teacher model. |
| distil_bert_base_multi | DistilBert | 134.73M | 6-layer DistilBERT model where case is maintained. Trained on Wikipedias of 104 languages |
| efficientnet_lite0_ra_imagenet | - | 4.65M | EfficientNet-Lite model fine-trained on the ImageNet 1k dataset with RandAugment recipe. |
| efficientnet_b0_ra_imagenet | - | 5.29M | EfficientNet B0 model pre-trained on the ImageNet 1k dataset with RandAugment recipe. |
| efficientnet_b0_ra4_e3600_r224_imagenet | - | 5.29M | EfficientNet B0 model pre-trained on the ImageNet 1k dataset by Ross Wightman. Trained with timm scripts using hyper-parameters inspired by the MobileNet-V4 small, mixed with go-to hparams from timm and "ResNet Strikes Back". |
| efficientnet_es_ra_imagenet | - | 5.44M | EfficientNet-EdgeTPU Small model trained on the ImageNet 1k dataset with RandAugment recipe. |
| efficientnet_em_ra2_imagenet | - | 6.90M | EfficientNet-EdgeTPU Medium model trained on the ImageNet 1k dataset with RandAugment2 recipe. |
| efficientnet_b1_ft_imagenet | - | 7.79M | EfficientNet B1 model fine-tuned on the ImageNet 1k dataset. |
| efficientnet_b1_ra4_e3600_r240_imagenet | - | 7.79M | EfficientNet B1 model pre-trained on the ImageNet 1k dataset by Ross Wightman. Trained with timm scripts using hyper-parameters inspired by the MobileNet-V4 small, mixed with go-to hparams from timm and "ResNet Strikes Back". |
| efficientnet_b2_ra_imagenet | - | 9.11M | EfficientNet B2 model pre-trained on the ImageNet 1k dataset with RandAugment recipe. |
| efficientnet_el_ra_imagenet | - | 10.59M | EfficientNet-EdgeTPU Large model trained on the ImageNet 1k dataset with RandAugment recipe. |
| efficientnet_b3_ra2_imagenet | - | 12.23M | EfficientNet B3 model pre-trained on the ImageNet 1k dataset with RandAugment2 recipe. |
| efficientnet_b4_ra2_imagenet | - | 19.34M | EfficientNet B4 model pre-trained on the ImageNet 1k dataset with RandAugment2 recipe. |
| efficientnet_b5_sw_imagenet | - | 30.39M | EfficientNet B5 model pre-trained on the ImageNet 12k dataset by Ross Wightman. Based on Swin Transformer train / pretrain recipe with modifications (related to both DeiT and ConvNeXt recipes). |
| efficientnet_b5_sw_ft_imagenet | - | 30.39M | EfficientNet B5 model pre-trained on the ImageNet 12k dataset and fine-tuned on ImageNet-1k by Ross Wightman. Based on Swin Transformer train / pretrain recipe with modifications (related to both DeiT and ConvNeXt recipes). |
| electra_small_discriminator_uncased_en | Electra | 13.55M | 12-layer small ELECTRA discriminator model. All inputs are lowercased. Trained on English Wikipedia + BooksCorpus. |
| electra_small_generator_uncased_en | Electra | 13.55M | 12-layer small ELECTRA generator model. All inputs are lowercased. Trained on English Wikipedia + BooksCorpus. |
| electra_base_generator_uncased_en | Electra | 33.58M | 12-layer base ELECTRA generator model. All inputs are lowercased. Trained on English Wikipedia + BooksCorpus. |
| electra_large_generator_uncased_en | Electra | 51.07M | 24-layer large ELECTRA generator model. All inputs are lowercased. Trained on English Wikipedia + BooksCorpus. |
| electra_base_discriminator_uncased_en | Electra | 109.48M | 12-layer base ELECTRA discriminator model. All inputs are lowercased. Trained on English Wikipedia + BooksCorpus. |
| electra_large_discriminator_uncased_en | Electra | 335.14M | 24-layer large ELECTRA discriminator model. All inputs are lowercased. Trained on English Wikipedia + BooksCorpus. |
| f_net_base_en | FNet | 82.86M | 12-layer FNet model where case is maintained. Trained on the C4 dataset. |
| f_net_large_en | FNet | 236.95M | 24-layer FNet model where case is maintained. Trained on the C4 dataset. |
| falcon_refinedweb_1b_en | Falcon | 1.31B | 24-layer Falcon model (Falcon with 1B parameters), trained on 350B tokens of RefinedWeb dataset. |
| schnell | - | 124.44M | A 12 billion parameter rectified flow transformer capable of generating images from text descriptions. |
| gemma_2b_en | Gemma | 2.51B | 2 billion parameter, 18-layer, base Gemma model. |
| gemma_instruct_2b_en | Gemma | 2.51B | 2 billion parameter, 18-layer, instruction tuned Gemma model. |
| gemma_1.1_instruct_2b_en | Gemma | 2.51B | 2 billion parameter, 18-layer, instruction tuned Gemma model. The 1.1 update improves model quality. |
| code_gemma_1.1_2b_en | Gemma | 2.51B | 2 billion parameter, 18-layer, CodeGemma model. This model has been trained on a fill-in-the-middle (FIM) task for code completion. The 1.1 update improves model quality. |
| code_gemma_2b_en | Gemma | 2.51B | 2 billion parameter, 18-layer, CodeGemma model. This model has been trained on a fill-in-the-middle (FIM) task for code completion. |
| gemma2_2b_en | Gemma | 2.61B | 2 billion parameter, 26-layer, base Gemma model. |
| gemma2_instruct_2b_en | Gemma | 2.61B | 2 billion parameter, 26-layer, instruction tuned Gemma model. |
| shieldgemma_2b_en | Gemma | 2.61B | 2 billion parameter, 26-layer, ShieldGemma model. |
| gemma_7b_en | Gemma | 8.54B | 7 billion parameter, 28-layer, base Gemma model. |
| gemma_instruct_7b_en | Gemma | 8.54B | 7 billion parameter, 28-layer, instruction tuned Gemma model. |
| gemma_1.1_instruct_7b_en | Gemma | 8.54B | 7 billion parameter, 28-layer, instruction tuned Gemma model. The 1.1 update improves model quality. |
| code_gemma_7b_en | Gemma | 8.54B | 7 billion parameter, 28-layer, CodeGemma model. This model has been trained on a fill-in-the-middle (FIM) task for code completion. |
| code_gemma_instruct_7b_en | Gemma | 8.54B | 7 billion parameter, 28-layer, instruction tuned CodeGemma model. This model has been trained for chat use cases related to code. |
| code_gemma_1.1_instruct_7b_en | Gemma | 8.54B | 7 billion parameter, 28-layer, instruction tuned CodeGemma model. This model has been trained for chat use cases related to code. The 1.1 update improves model quality. |
| gemma2_9b_en | Gemma | 9.24B | 9 billion parameter, 42-layer, base Gemma model. |
| gemma2_instruct_9b_en | Gemma | 9.24B | 9 billion parameter, 42-layer, instruction tuned Gemma model. |
| shieldgemma_9b_en | Gemma | 9.24B | 9 billion parameter, 42-layer, ShieldGemma model. |
| gemma2_27b_en | Gemma | 27.23B | 27 billion parameter, 42-layer, base Gemma model. |
| gemma2_instruct_27b_en | Gemma | 27.23B | 27 billion parameter, 42-layer, instruction tuned Gemma model. |
| shieldgemma_27b_en | Gemma | 27.23B | 27 billion parameter, 42-layer, ShieldGemma model. |
| gpt2_base_en | GPT2 | 124.44M | 12-layer GPT-2 model where case is maintained. Trained on WebText. |
| gpt2_base_en_cnn_dailymail | GPT2 | 124.44M | 12-layer GPT-2 model where case is maintained. Finetuned on the CNN/DailyMail summarization dataset. |
| gpt2_medium_en | GPT2 | 354.82M | 24-layer GPT-2 model where case is maintained. Trained on WebText. |
| gpt2_large_en | GPT2 | 774.03M | 36-layer GPT-2 model where case is maintained. Trained on WebText. |
| gpt2_extra_large_en | GPT2 | 1.56B | 48-layer GPT-2 model where case is maintained. Trained on WebText. |
| llama2_7b_en | Llama | 6.74B | 7 billion parameter, 32-layer, base LLaMA 2 model. |
| llama2_instruct_7b_en | Llama | 6.74B | 7 billion parameter, 32-layer, instruction tuned LLaMA 2 model. |
| vicuna_1.5_7b_en | Llama | 6.74B | 7 billion parameter, 32-layer, instruction tuned Vicuna v1.5 model. |
| llama2_7b_en_int8 | Llama | 6.74B | 7 billion parameter, 32-layer, base LLaMA 2 model with activation and weights quantized to int8. |
| llama2_instruct_7b_en_int8 | Llama | 6.74B | 7 billion parameter, 32-layer, instruction tuned LLaMA 2 model with activation and weights quantized to int8. |
| llama3_8b_en | Llama3 | 8.03B | 8 billion parameter, 32-layer, base LLaMA 3 model. |
| llama3_instruct_8b_en | Llama3 | 8.03B | 8 billion parameter, 32-layer, instruction tuned LLaMA 3 model. |
| llama3_8b_en_int8 | Llama3 | 8.03B | 8 billion parameter, 32-layer, base LLaMA 3 model with activation and weights quantized to int8. |
| llama3_instruct_8b_en_int8 | Llama3 | 8.03B | 8 billion parameter, 32-layer, instruction tuned LLaMA 3 model with activation and weights quantized to int8. |
| mistral_7b_en | Mistral | 7.24B | Mistral 7B base model |
| mistral_instruct_7b_en | Mistral | 7.24B | Mistral 7B instruct model |
| mistral_0.2_instruct_7b_en | Mistral | 7.24B | Mistral 7B instruct Version 0.2 model |
| mit_b0_ade20k_512 | MiT | 3.32M | MiT (MixTransformer) model with 8 transformer blocks. |
| mit_b0_cityscapes_1024 | MiT | 3.32M | MiT (MixTransformer) model with 8 transformer blocks. |
| mit_b1_ade20k_512 | MiT | 13.16M | MiT (MixTransformer) model with 8 transformer blocks. |
| mit_b1_cityscapes_1024 | MiT | 13.16M | MiT (MixTransformer) model with 8 transformer blocks. |
| mit_b2_ade20k_512 | MiT | 24.20M | MiT (MixTransformer) model with 16 transformer blocks. |
| mit_b2_cityscapes_1024 | MiT | 24.20M | MiT (MixTransformer) model with 16 transformer blocks. |
| mit_b3_ade20k_512 | MiT | 44.08M | MiT (MixTransformer) model with 28 transformer blocks. |
| mit_b3_cityscapes_1024 | MiT | 44.08M | MiT (MixTransformer) model with 28 transformer blocks. |
| mit_b4_ade20k_512 | MiT | 60.85M | MiT (MixTransformer) model with 41 transformer blocks. |
| mit_b4_cityscapes_1024 | MiT | 60.85M | MiT (MixTransformer) model with 41 transformer blocks. |
| mit_b5_ade20k_640 | MiT | 81.45M | MiT (MixTransformer) model with 52 transformer blocks. |
| mit_b5_cityscapes_1024 | MiT | 81.45M | MiT (MixTransformer) model with 52 transformer blocks. |
| opt_125m_en | OPT | 125.24M | 12-layer OPT model where case in maintained. Trained on BookCorpus, CommonCrawl, Pile, and PushShift.io corpora. |
| opt_1.3b_en | OPT | 1.32B | 24-layer OPT model where case in maintained. Trained on BookCorpus, CommonCrawl, Pile, and PushShift.io corpora. |
| opt_2.7b_en | OPT | 2.70B | 32-layer OPT model where case in maintained. Trained on BookCorpus, CommonCrawl, Pile, and PushShift.io corpora. |
| opt_6.7b_en | OPT | 6.70B | 32-layer OPT model where case in maintained. Trained on BookCorpus, CommonCrawl, Pile, and PushShift.io corpora. |
| pali_gemma_3b_mix_224 | PaliGemma | 2.92B | image size 224, mix fine tuned, text sequence length is 256 |
| pali_gemma_3b_224 | PaliGemma | 2.92B | image size 224, pre trained, text sequence length is 128 |
| pali_gemma_3b_mix_448 | PaliGemma | 2.92B | image size 448, mix fine tuned, text sequence length is 512 |
| pali_gemma_3b_448 | PaliGemma | 2.92B | image size 448, pre trained, text sequence length is 512 |
| pali_gemma_3b_896 | PaliGemma | 2.93B | image size 896, pre trained, text sequence length is 512 |
| pali_gemma2_pt_3b_224 | - | 3.03B | 3 billion parameter, image size 224, 27-layer for SigLIP-So400m vision encoder and 26-layer Gemma2 2B lanuage model. This model has been pre-trained on a mixture of datasets. |
| pali_gemma2_3b_ft_docci_448 | - | 3.03B | 3 billion parameter, image size 448, 27-layer for SigLIP-So400m vision encoder and 26-layer Gemma2 2B lanuage model. This model has been fine-tuned on the DOCCI dataset for improved descriptions with fine-grained details. |
| pali_gemma2_pt_3b_448 | - | 3.03B | 3 billion parameter, image size 448, 27-layer for SigLIP-So400m vision encoder and 26-layer Gemma2 2B lanuage model. This model has been pre-trained on a mixture of datasets. |
| pali_gemma2_pt_3b_896 | - | 3.04B | 3 billion parameter, image size 896, 27-layer for SigLIP-So400m vision encoder and 26-layer Gemma2 2B lanuage model. This model has been pre-trained on a mixture of datasets. |
| pali_gemma2_pt_10b_224 | - | 9.66B | 10 billion parameter, image size 224, 27-layer for SigLIP-So400m vision encoder and 42-layer Gemma2 9B lanuage model. This model has been pre-trained on a mixture of datasets. |
| pali_gemma2_pt_28b_224 | - | 9.66B | 28 billion parameter, image size 224, 27-layer for SigLIP-So400m vision encoder and 46-layer Gemma2 27B lanuage model. This model has been pre-trained on a mixture of datasets. |
| pali_gemma2_10b_ft_docci_448 | - | 9.66B | 10 billion parameter, 27-layer for SigLIP-So400m vision encoder and 42-layer Gemma2 9B lanuage model. This model has been fine-tuned on the DOCCI dataset for improved descriptions with fine-grained details. |
| pali_gemma2_pt_10b_448 | - | 9.66B | 10 billion parameter, image size 448, 27-layer for SigLIP-So400m vision encoder and 42-layer Gemma2 9B lanuage model. This model has been pre-trained on a mixture of datasets. |
| pali_gemma2_pt_28b_448 | - | 9.66B | 28 billion parameter, image size 448, 27-layer for SigLIP-So400m vision encoder and 46-layer Gemma2 27B lanuage model. This model has been pre-trained on a mixture of datasets. |
| pali_gemma2_pt_10b_896 | - | 9.67B | 10 billion parameter, image size 896, 27-layer for SigLIP-So400m vision encoder and 42-layer Gemma2 9B lanuage model. This model has been pre-trained on a mixture of datasets. |
| pali_gemma2_pt_28b_896 | - | 9.67B | 28 billion parameter, image size 896, 27-layer for SigLIP-So400m vision encoder and 46-layer Gemma2 27B lanuage model. This model has been pre-trained on a mixture of datasets. |
| phi3_mini_4k_instruct_en | Phi3 | 3.82B | 3.8 billion parameters, 32 layers, 4k context length, Phi-3 model. The model was trained using the Phi-3 datasets. This dataset includes both synthetic data and filtered publicly available website data, with an emphasis on high-quality and reasoning-dense properties. |
| phi3_mini_128k_instruct_en | Phi3 | 3.82B | 3.8 billion parameters, 32 layers, 128k context length, Phi-3 model. The model was trained using the Phi-3 datasets. This dataset includes both synthetic data and filtered publicly available website data, with an emphasis on high-quality and reasoning-dense properties. |
| resnet_18_imagenet | ResNet | 11.19M | 18-layer ResNet model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_vd_18_imagenet | ResNet | 11.72M | 18-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_vd_34_imagenet | ResNet | 21.84M | 34-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_50_imagenet | ResNet | 23.56M | 50-layer ResNet model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_v2_50_imagenet | ResNet | 23.56M | 50-layer ResNetV2 model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_vd_50_imagenet | ResNet | 25.63M | 50-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_vd_50_ssld_imagenet | ResNet | 25.63M | 50-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution with knowledge distillation. |
| resnet_vd_50_ssld_v2_imagenet | ResNet | 25.63M | 50-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution with knowledge distillation and AutoAugment. |
| resnet_vd_50_ssld_v2_fix_imagenet | ResNet | 25.63M | 50-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution with knowledge distillation, AutoAugment and additional fine-tuning of the classification head. |
| resnet_101_imagenet | ResNet | 42.61M | 101-layer ResNet model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_v2_101_imagenet | ResNet | 42.61M | 101-layer ResNetV2 model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_vd_101_imagenet | ResNet | 44.67M | 101-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_vd_101_ssld_imagenet | ResNet | 44.67M | 101-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution with knowledge distillation. |
| resnet_152_imagenet | ResNet | 58.30M | 152-layer ResNet model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_vd_152_imagenet | ResNet | 60.36M | 152-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| resnet_vd_200_imagenet | ResNet | 74.93M | 200-layer ResNetVD (ResNet with bag of tricks) model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| retinanet_resnet50_fpn_coco | - | 34.12M | RetinaNet model with ResNet50 backbone fine-tuned on COCO in 800x800 resolution. |
| roberta_base_en | Roberta | 124.05M | 12-layer RoBERTa model where case is maintained.Trained on English Wikipedia, BooksCorpus, CommonCraw, and OpenWebText. |
| roberta_large_en | Roberta | 354.31M | 24-layer RoBERTa model where case is maintained.Trained on English Wikipedia, BooksCorpus, CommonCraw, and OpenWebText. |
| sam_base_sa1b | Segment Anything Model | 93.74M | The base SAM model trained on the SA1B dataset. |
| sam_huge_sa1b | Segment Anything Model | 312.34M | The huge SAM model trained on the SA1B dataset. |
| sam_large_sa1b | Segment Anything Model | 641.09M | The large SAM model trained on the SA1B dataset. |
| stable_diffusion_3_medium | Stable Diffusion 3 | 2.99B | 3 billion parameter, including CLIP L and CLIP G text encoders, MMDiT generative model, and VAE autoencoder. Developed by Stability AI. |
| stable_diffusion_3.5_large | Stable Diffusion 3 | 9.05B | 9 billion parameter, including CLIP L and CLIP G text encoders, MMDiT generative model, and VAE autoencoder. Developed by Stability AI. |
| stable_diffusion_3.5_large_turbo | Stable Diffusion 3 | 9.05B | 9 billion parameter, including CLIP L and CLIP G text encoders, MMDiT generative model, and VAE autoencoder. A timestep-distilled version that eliminates classifier-free guidance and uses fewer steps for generation. Developed by Stability AI. |
| t5_small_multi | T5 | 0 | 8-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| t5_base_multi | T5 | 0 | 12-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| t5_large_multi | T5 | 0 | 24-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| flan_small_multi | T5 | 0 | 8-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| flan_base_multi | T5 | 0 | 12-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| flan_large_multi | T5 | 0 | 24-layer T5 model. Trained on the Colossal Clean Crawled Corpus (C4). |
| t5_1.1_small | T5 | 60.51M | |
| t5_1.1_base | T5 | 247.58M | |
| t5_1.1_large | T5 | 750.25M | |
| t5_1.1_xl | T5 | 2.85B | |
| t5_1.1_xxl | T5 | 11.14B | |
| vgg_11_imagenet | VGG | 9.22M | 11-layer vgg model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| vgg_13_imagenet | VGG | 9.40M | 13-layer vgg model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| vgg_16_imagenet | VGG | 14.71M | 16-layer vgg model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| vgg_19_imagenet | VGG | 20.02M | 19-layer vgg model pre-trained on the ImageNet 1k dataset at a 224x224 resolution. |
| whisper_tiny_en | Whisper | 37.18M | 4-layer Whisper model. Trained on 438,000 hours of labelled English speech data. |
| whisper_tiny_multi | Whisper | 37.76M | 4-layer Whisper model. Trained on 680,000 hours of labelled multilingual speech data. |
| whisper_base_multi | Whisper | 72.59M | 6-layer Whisper model. Trained on 680,000 hours of labelled multilingual speech data. |
| whisper_base_en | Whisper | 124.44M | 6-layer Whisper model. Trained on 438,000 hours of labelled English speech data. |
| whisper_small_en | Whisper | 241.73M | 12-layer Whisper model. Trained on 438,000 hours of labelled English speech data. |
| whisper_small_multi | Whisper | 241.73M | 12-layer Whisper model. Trained on 680,000 hours of labelled multilingual speech data. |
| whisper_medium_en | Whisper | 763.86M | 24-layer Whisper model. Trained on 438,000 hours of labelled English speech data. |
| whisper_medium_multi | Whisper | 763.86M | 24-layer Whisper model. Trained on 680,000 hours of labelled multilingual speech data. |
| whisper_large_multi | Whisper | 1.54B | 32-layer Whisper model. Trained on 680,000 hours of labelled multilingual speech data. |
| whisper_large_multi_v2 | Whisper | 1.54B | 32-layer Whisper model. Trained for 2.5 epochs on 680,000 hours of labelled multilingual speech data. An improved of whisper_large_multi. |
| xlm_roberta_base_multi | XLMRoberta | 277.45M | 12-layer XLM-RoBERTa model where case is maintained. Trained on CommonCrawl in 100 languages. |
| xlm_roberta_large_multi | XLMRoberta | 558.84M | 24-layer XLM-RoBERTa model where case is maintained. Trained on CommonCrawl in 100 languages. |