AI & ML interests
Open-Source ML
Recent Activity
Stable Diffusion 3.5 large, large-turbo, medium and the shared T5-XXL encoder as pure Keras 3 models (zeromodels), JAX / PyTorch / TensorFlow.
SDXL base, refiner and SDXL-Turbo as pure Keras 3 models (zeromodels): one container per checkpoint, runs on JAX, PyTorch and TensorFlow.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) LeViT image-classification checkpoints - a hybrid conv/transformer ImageNet classifier for fast inference.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) Table Transformer table detection and table-structure recognition checkpoints
-
zeromodels/table-transformer-detection
Object Detection • Updated • 24 -
zeromodels/table-transformer-structure-recognition
Object Detection • Updated • 21 -
zeromodels/table-transformer-structure-recognition-v1.1-all
Object Detection • Updated • 43 -
zeromodels/table-transformer-structure-recognition-v1.1-fin
Object Detection • Updated • 21
Pyramid Vision Transformer v1 and v2 image classifiers, converted to zeromodels (Keras 3). ImageNet-1k, load with from_weights.
-
zeromodels/owlv2-large-patch14-finetuned
Zero-Shot Object Detection • Updated • 19 -
zeromodels/owlv2-large-patch14-ensemble
Zero-Shot Object Detection • Updated • 89 -
zeromodels/owlv2-large-patch14
Zero-Shot Object Detection • Updated • 25 -
zeromodels/owlv2-base-patch16-finetuned
Zero-Shot Object Detection • Updated • 20
-
zeromodels/mask2former-swin-tiny-coco-instance
Image Segmentation • Updated • 29 -
zeromodels/mask2former-swin-base-coco-instance
Image Segmentation • Updated • 47 -
zeromodels/mask2former-swin-tiny-ade-semantic
Image Segmentation • Updated • 59 -
zeromodels/mask2former-swin-tiny-coco-panoptic
Image Segmentation • Updated • 48
Pure-Keras 3 CLIP weights for kerasformers.
-
zeromodels/clip_vit_base_16
Zero-Shot Image Classification • Updated • 49 -
zeromodels/clip_vit_base_32
Zero-Shot Image Classification • Updated • 35 -
zeromodels/clip_vit_large_14
Zero-Shot Image Classification • Updated • 28 -
zeromodels/clip_vit_large_14_336
Zero-Shot Image Classification • Updated • 33
Pure-Keras 3 SigLIP weights for kerasformers.
-
zeromodels/siglip_base_p16_224
Zero-Shot Image Classification • Updated • 31 -
zeromodels/siglip_base_p16_256
Zero-Shot Image Classification • Updated • 22 -
zeromodels/siglip_base_p16_multilingual_256
Zero-Shot Image Classification • Updated • 31 -
zeromodels/siglip_base_p16_384
Zero-Shot Image Classification • Updated • 33
Pure-Keras 3 LocateAnything weights for kerasformers.
Pure-Keras 3 Speech2Text weights for kerasformers.
Pure-Keras 3 Granite Speech weights for kerasformers.
-
zeromodels/granite_speech_3_3_2b
Automatic Speech Recognition • Updated • 25 • 1 -
zeromodels/granite_speech_3_3_8b
Automatic Speech Recognition • Updated • 26 -
zeromodels/granite_speech_4_1_2b
Automatic Speech Recognition • Updated • 18 -
zeromodels/granite_4_0_1b_speech
Automatic Speech Recognition • Updated • 19
Pure-Keras 3 Gemma 2 weights for kerasformers.
Pure-Keras 3 ConvMixer image-classification weights for kerasformers.
Pure-Keras 3 ConvNeXt-V2 image-classification weights for kerasformers.
-
zeromodels/convnextv2_atto_fcmae_ft_in1k
Image Classification • Updated • 46 -
zeromodels/convnextv2_base_fcmae_ft_in1k
Image Classification • Updated • 45 -
zeromodels/convnextv2_base_fcmae_ft_in22k_in1k
Image Classification • Updated • 49 -
zeromodels/convnextv2_base_fcmae_ft_in22k_in1k_384
Image Classification • Updated • 40
Pure-Keras 3 DenseNet image-classification weights for kerasformers.
Pure-Keras 3 FlexiViT image-classification weights for kerasformers.
Pure-Keras 3 Inception-ResNet-V2 image-classification weights for kerasformers.
Pure-Keras 3 Inception-V4 image-classification weights for kerasformers.
Pure-Keras 3 MLP-Mixer image-classification weights for kerasformers.
-
zeromodels/mixer_b16_224_goog_in21k
Image Classification • Updated • 64 -
zeromodels/mixer_b16_224_goog_in21k_ft_in1k
Image Classification • Updated • 39 -
zeromodels/mixer_b16_224_miil_in21k_ft_in1k
Image Classification • Updated • 46 -
zeromodels/mixer_l16_224_goog_in21k
Image Classification • Updated • 43
Pure-Keras 3 MobileNetV3 image-classification weights for kerasformers.
-
zeromodels/mobilenetv3_large_100_miil_in21k
Image Classification • Updated • 27 -
zeromodels/mobilenetv3_large_100_miil_in21k_ft_in1k
Image Classification • Updated • 34 -
zeromodels/mobilenetv3_large_100_ra4_e3600_r224_in1k
Image Classification • Updated • 41 -
zeromodels/mobilenetv3_large_100_ra_in1k
Image Classification • Updated • 31
Pure-Keras 3 PiT image-classification weights for kerasformers.
Pure-Keras 3 Res2Net image-classification weights for kerasformers.
Pure-Keras 3 ResNet image-classification weights for kerasformers.
Pure-Keras 3 ResNeXt image-classification weights for kerasformers.
-
zeromodels/resnext101_32x16d_fb_ssl_yfcc100m_ft_in1k
Image Classification • Updated • 54 -
zeromodels/resnext101_32x16d_fb_swsl_ig1b_ft_in1k
Image Classification • Updated • 39 -
zeromodels/resnext101_32x16d_fb_wsl_ig1b_ft_in1k
Image Classification • Updated • 49 -
zeromodels/resnext101_32x32d_fb_wsl_ig1b_ft_in1k
Image Classification • Updated • 49
Pure-Keras 3 Swin Transformer image-classification weights for kerasformers.
-
zeromodels/swin_base_patch4_window12_384_ms_in1k
Image Classification • Updated • 44 -
zeromodels/swin_base_patch4_window12_384_ms_in22k
Image Classification • Updated • 49 -
zeromodels/swin_base_patch4_window12_384_ms_in22k_ft_in1k
Image Classification • Updated • 51 -
zeromodels/swin_base_patch4_window7_224_ms_in1k
Image Classification • Updated • 42
Pure-Keras 3 EfficientNet image-classification weights for kerasformers.
Pure-Keras 3 EfficientNetV2 image-classification weights for kerasformers.
Pure-Keras 3 ViT image-classification weights for kerasformers.
-
zeromodels/vit_base_patch16_224_augreg_in1k
Image Classification • Updated • 39 -
zeromodels/vit_base_patch16_224_augreg_in21k
Image Classification • Updated • 46 -
zeromodels/vit_base_patch16_224_augreg_in21k_ft_in1k
Image Classification • Updated • 52 -
zeromodels/vit_base_patch16_224_orig_in21k_ft_in1k
Image Classification • Updated • 43
Pure-Keras 3 MiT (SegFormer encoder) weights (classification and segmentation) for kerasformers.
Pure-Keras 3 MobileViT-V2 weights (classification and segmentation) for kerasformers.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of BERT encoders and masked-LM heads.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of DeBERTa, DeBERTa-v2, and DeBERTa-v3 encoders.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the DeepSeek Janus-Pro any-to-any multimodal models.
OpenAI's open-weight GPT-OSS mixture-of-experts LLMs converted to pure Keras 3 for kerasformers, experts kept in MXFP4.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Google's Gemma 4, for kerasformers.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Qwen 2.5 VL for kerasformers.
Pure-Keras 3 conversions of Qwen3.5-MoE (kerasformers).
Pure-Keras 3 conversions of Qwen2-MoE sparse Mixture-of-Experts LLMs (incl. Qwen1.5-MoE) (kerasformers).
Pure-Keras 3 conversions of Qwen3.8 (kerasformers).
Pure-Keras 3 conversions of GLM (kerasformers).
Pure-Keras 3 (JAX / PyTorch / TensorFlow) MobileNetV4 checkpoints for kerasformers, converted from timm.
-
zeromodels/mobilenetv4_conv_small_e2400_r224_in1k
Image Classification • Updated • 38 -
zeromodels/mobilenetv4_conv_medium_e500_r256_in1k
Image Classification • Updated • 39 -
zeromodels/mobilenetv4_conv_large_e600_r384_in1k
Image Classification • Updated • 29 -
zeromodels/mobilenetv4_hybrid_medium_e500_r224_in1k
Image Classification • Updated • 36
Pure-Keras 3 (JAX/PyTorch/TF) TIPSv2-DPT depth + segmentation models converted from google/tipsv2-dpt
MPNet in pure Keras 3 - runs unmodified on JAX, PyTorch and TensorFlow.
Stable Diffusion 3 medium and its shared T5-XXL text encoder as pure Keras 3 models (zeromodels), runs on JAX, PyTorch and TensorFlow.
Stable Diffusion 2.x and SD-Turbo as pure Keras 3 models (zeromodels): one container per checkpoint, runs on JAX, PyTorch and TensorFlow.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) BEiT image-classification and semantic-segmentation checkpoints.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) RegNet X and Y backbones and classifiers, converted from the official facebook/regnet-* checkpoints.
Family of scalable and efficient object detection models developed by Google Research
Pure-Keras 3 MetaCLIP 2 weights for kerasformers.
-
zeromodels/metaclip2_worldwide_s16_224
Zero-Shot Image Classification • Updated • 32 -
zeromodels/metaclip2_worldwide_s16_384
Zero-Shot Image Classification • Updated • 32 -
zeromodels/metaclip2_worldwide_m16_224
Zero-Shot Image Classification • Updated • 23 -
zeromodels/metaclip2_worldwide_m16_384
Zero-Shot Image Classification • Updated • 16
Pure-Keras 3 SigLIP2 weights for kerasformers.
-
zeromodels/siglip2_base_p16_224
Zero-Shot Image Classification • Updated • 26 -
zeromodels/siglip2_base_p16_256
Zero-Shot Image Classification • Updated • 29 -
zeromodels/siglip2_base_p16_384
Zero-Shot Image Classification • Updated • 6.86k -
zeromodels/siglip2_base_p16_512
Zero-Shot Image Classification • Updated • 36
Pure-Keras 3 Whisper weights for kerasformers.
Pure-Keras 3 Moonshine weights for kerasformers.
Pure-Keras 3 Granite Speech Plus weights for kerasformers.
Pure-Keras 3 CaiT image-classification weights for kerasformers.
Pure-Keras 3 ConvNeXt image-classification weights for kerasformers.
Pure-Keras 3 DeiT and DeiT3 image-classification weights for kerasformers.
-
zeromodels/deit3_base_patch16_224_fb_in1k
Image Classification • Updated • 33 -
zeromodels/deit3_base_patch16_224_fb_in22k_ft_in1k
Image Classification • Updated • 46 -
zeromodels/deit3_base_patch16_384_fb_in1k
Image Classification • Updated • 40 -
zeromodels/deit3_base_patch16_384_fb_in22k_ft_in1k
Image Classification • Updated • 34
Pure-Keras 3 EfficientFormer image-classification weights for kerasformers.
Pure-Keras 3 InceptionNeXt image-classification weights for kerasformers.
-
zeromodels/inception_next_atto_sail_in1k
Image Classification • Updated • 57 -
zeromodels/inception_next_base_sail_in1k
Image Classification • Updated • 41 -
zeromodels/inception_next_base_sail_in1k_384
Image Classification • Updated • 53 -
zeromodels/inception_next_small_sail_in1k
Image Classification • Updated • 59
Pure-Keras 3 Inception-V3 image-classification weights for kerasformers.
Pure-Keras 3 MaxViT image-classification weights for kerasformers.
Pure-Keras 3 MobileNetV2 image-classification weights for kerasformers.
Pure-Keras 3 Next-ViT image-classification weights for kerasformers.
Pure-Keras 3 PoolFormer image-classification weights for kerasformers.
Pure-Keras 3 ResMLP image-classification weights for kerasformers.
Pure-Keras 3 ResNetV2 (BiT) image-classification weights for kerasformers.
-
zeromodels/resnetv2_101x1_bit_goog_in21k
Image Classification • Updated • 38 -
zeromodels/resnetv2_101x1_bit_goog_in21k_ft_in1k
Image Classification • Updated • 44 -
zeromodels/resnetv2_101x3_bit_goog_in21k
Image Classification • Updated • 580 -
zeromodels/resnetv2_101x3_bit_goog_in21k_ft_in1k
Image Classification • Updated • 44
Pure-Keras 3 SE-ResNet and SE-ResNeXt image-classification weights for kerasformers.
Pure-Keras 3 Swin Transformer V2 image-classification weights for kerasformers.
-
zeromodels/swinv2_base_window12_192_ms_in22k
Image Classification • Updated • 44 -
zeromodels/swinv2_base_window12to16_192to256_ms_in22k_ft_in1k
Image Classification • Updated • 41 -
zeromodels/swinv2_base_window12to24_192to384_ms_in22k_ft_in1k
Image Classification • Updated • 33 -
zeromodels/swinv2_base_window16_256_ms_in1k
Image Classification • Updated • 30
Pure-Keras 3 EfficientNet-Lite image-classification weights for kerasformers.
Pure-Keras 3 VGG image-classification weights for kerasformers.
Pure-Keras 3 Xception image-classification weights for kerasformers.
Pure-Keras 3 MobileViT weights (classification and segmentation) for kerasformers.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of RoBERTa encoders and masked-LM heads.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the multilingual XLM-RoBERTa encoders and masked-LM heads.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the DeepSeek-VL vision-language models.
Pure-Keras 3 GPT and GPT-2 language models for kerasformers.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Google's Gemma 3, for kerasformers.
kerasformers ports of Google's Gemma 3n on-device multimodal models (image + audio + text), runnable on JAX / PyTorch / TensorFlow.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Qwen 2 VL for kerasformers.
Pure-Keras 3 conversions of Qwen3-Next-80B-A3B (kerasformers).
Pure-Keras 3 conversions of Qwen3-VL-MoE (kerasformers).
Pure-Keras 3 conversions of Qwen3-MoE sparse Mixture-of-Experts LLMs (kerasformers).
Pure-Keras 3 conversions of InternVL 3 and 3.5 (kerasformers).
Pure-Keras 3 conversions of the original T5 (kerasformers).
Google's ELECTRA text encoders in pure Keras 3 (kerasformers), runnable on JAX / PyTorch / TensorFlow.
Pure-Keras 3 (JAX/PyTorch/TF) TIPSv2 vision-language dual encoders, converted from google/tipsv2
MPNet in pure Keras 3 - runs unmodified on JAX, PyTorch and TensorFlow.
Stable Diffusion 3.5 large, large-turbo, medium and the shared T5-XXL encoder as pure Keras 3 models (zeromodels), JAX / PyTorch / TensorFlow.
Stable Diffusion 3 medium and its shared T5-XXL text encoder as pure Keras 3 models (zeromodels), runs on JAX, PyTorch and TensorFlow.
SDXL base, refiner and SDXL-Turbo as pure Keras 3 models (zeromodels): one container per checkpoint, runs on JAX, PyTorch and TensorFlow.
Stable Diffusion 2.x and SD-Turbo as pure Keras 3 models (zeromodels): one container per checkpoint, runs on JAX, PyTorch and TensorFlow.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) LeViT image-classification checkpoints - a hybrid conv/transformer ImageNet classifier for fast inference.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) BEiT image-classification and semantic-segmentation checkpoints.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) Table Transformer table detection and table-structure recognition checkpoints
-
zeromodels/table-transformer-detection
Object Detection • Updated • 24 -
zeromodels/table-transformer-structure-recognition
Object Detection • Updated • 21 -
zeromodels/table-transformer-structure-recognition-v1.1-all
Object Detection • Updated • 43 -
zeromodels/table-transformer-structure-recognition-v1.1-fin
Object Detection • Updated • 21
Pure-Keras 3 (JAX / PyTorch / TensorFlow) RegNet X and Y backbones and classifiers, converted from the official facebook/regnet-* checkpoints.
Pyramid Vision Transformer v1 and v2 image classifiers, converted to zeromodels (Keras 3). ImageNet-1k, load with from_weights.
Family of scalable and efficient object detection models developed by Google Research
-
zeromodels/owlv2-large-patch14-finetuned
Zero-Shot Object Detection • Updated • 19 -
zeromodels/owlv2-large-patch14-ensemble
Zero-Shot Object Detection • Updated • 89 -
zeromodels/owlv2-large-patch14
Zero-Shot Object Detection • Updated • 25 -
zeromodels/owlv2-base-patch16-finetuned
Zero-Shot Object Detection • Updated • 20
-
zeromodels/mask2former-swin-tiny-coco-instance
Image Segmentation • Updated • 29 -
zeromodels/mask2former-swin-base-coco-instance
Image Segmentation • Updated • 47 -
zeromodels/mask2former-swin-tiny-ade-semantic
Image Segmentation • Updated • 59 -
zeromodels/mask2former-swin-tiny-coco-panoptic
Image Segmentation • Updated • 48
Pure-Keras 3 CLIP weights for kerasformers.
-
zeromodels/clip_vit_base_16
Zero-Shot Image Classification • Updated • 49 -
zeromodels/clip_vit_base_32
Zero-Shot Image Classification • Updated • 35 -
zeromodels/clip_vit_large_14
Zero-Shot Image Classification • Updated • 28 -
zeromodels/clip_vit_large_14_336
Zero-Shot Image Classification • Updated • 33
Pure-Keras 3 MetaCLIP 2 weights for kerasformers.
-
zeromodels/metaclip2_worldwide_s16_224
Zero-Shot Image Classification • Updated • 32 -
zeromodels/metaclip2_worldwide_s16_384
Zero-Shot Image Classification • Updated • 32 -
zeromodels/metaclip2_worldwide_m16_224
Zero-Shot Image Classification • Updated • 23 -
zeromodels/metaclip2_worldwide_m16_384
Zero-Shot Image Classification • Updated • 16
Pure-Keras 3 SigLIP weights for kerasformers.
-
zeromodels/siglip_base_p16_224
Zero-Shot Image Classification • Updated • 31 -
zeromodels/siglip_base_p16_256
Zero-Shot Image Classification • Updated • 22 -
zeromodels/siglip_base_p16_multilingual_256
Zero-Shot Image Classification • Updated • 31 -
zeromodels/siglip_base_p16_384
Zero-Shot Image Classification • Updated • 33
Pure-Keras 3 SigLIP2 weights for kerasformers.
-
zeromodels/siglip2_base_p16_224
Zero-Shot Image Classification • Updated • 26 -
zeromodels/siglip2_base_p16_256
Zero-Shot Image Classification • Updated • 29 -
zeromodels/siglip2_base_p16_384
Zero-Shot Image Classification • Updated • 6.86k -
zeromodels/siglip2_base_p16_512
Zero-Shot Image Classification • Updated • 36
Pure-Keras 3 LocateAnything weights for kerasformers.
Pure-Keras 3 Whisper weights for kerasformers.
Pure-Keras 3 Speech2Text weights for kerasformers.
Pure-Keras 3 Moonshine weights for kerasformers.
Pure-Keras 3 Granite Speech weights for kerasformers.
-
zeromodels/granite_speech_3_3_2b
Automatic Speech Recognition • Updated • 25 • 1 -
zeromodels/granite_speech_3_3_8b
Automatic Speech Recognition • Updated • 26 -
zeromodels/granite_speech_4_1_2b
Automatic Speech Recognition • Updated • 18 -
zeromodels/granite_4_0_1b_speech
Automatic Speech Recognition • Updated • 19
Pure-Keras 3 Granite Speech Plus weights for kerasformers.
Pure-Keras 3 Gemma 2 weights for kerasformers.
Pure-Keras 3 CaiT image-classification weights for kerasformers.
Pure-Keras 3 ConvMixer image-classification weights for kerasformers.
Pure-Keras 3 ConvNeXt image-classification weights for kerasformers.
Pure-Keras 3 ConvNeXt-V2 image-classification weights for kerasformers.
-
zeromodels/convnextv2_atto_fcmae_ft_in1k
Image Classification • Updated • 46 -
zeromodels/convnextv2_base_fcmae_ft_in1k
Image Classification • Updated • 45 -
zeromodels/convnextv2_base_fcmae_ft_in22k_in1k
Image Classification • Updated • 49 -
zeromodels/convnextv2_base_fcmae_ft_in22k_in1k_384
Image Classification • Updated • 40
Pure-Keras 3 DeiT and DeiT3 image-classification weights for kerasformers.
-
zeromodels/deit3_base_patch16_224_fb_in1k
Image Classification • Updated • 33 -
zeromodels/deit3_base_patch16_224_fb_in22k_ft_in1k
Image Classification • Updated • 46 -
zeromodels/deit3_base_patch16_384_fb_in1k
Image Classification • Updated • 40 -
zeromodels/deit3_base_patch16_384_fb_in22k_ft_in1k
Image Classification • Updated • 34
Pure-Keras 3 DenseNet image-classification weights for kerasformers.
Pure-Keras 3 EfficientFormer image-classification weights for kerasformers.
Pure-Keras 3 FlexiViT image-classification weights for kerasformers.
Pure-Keras 3 InceptionNeXt image-classification weights for kerasformers.
-
zeromodels/inception_next_atto_sail_in1k
Image Classification • Updated • 57 -
zeromodels/inception_next_base_sail_in1k
Image Classification • Updated • 41 -
zeromodels/inception_next_base_sail_in1k_384
Image Classification • Updated • 53 -
zeromodels/inception_next_small_sail_in1k
Image Classification • Updated • 59
Pure-Keras 3 Inception-ResNet-V2 image-classification weights for kerasformers.
Pure-Keras 3 Inception-V3 image-classification weights for kerasformers.
Pure-Keras 3 Inception-V4 image-classification weights for kerasformers.
Pure-Keras 3 MaxViT image-classification weights for kerasformers.
Pure-Keras 3 MLP-Mixer image-classification weights for kerasformers.
-
zeromodels/mixer_b16_224_goog_in21k
Image Classification • Updated • 64 -
zeromodels/mixer_b16_224_goog_in21k_ft_in1k
Image Classification • Updated • 39 -
zeromodels/mixer_b16_224_miil_in21k_ft_in1k
Image Classification • Updated • 46 -
zeromodels/mixer_l16_224_goog_in21k
Image Classification • Updated • 43
Pure-Keras 3 MobileNetV2 image-classification weights for kerasformers.
Pure-Keras 3 MobileNetV3 image-classification weights for kerasformers.
-
zeromodels/mobilenetv3_large_100_miil_in21k
Image Classification • Updated • 27 -
zeromodels/mobilenetv3_large_100_miil_in21k_ft_in1k
Image Classification • Updated • 34 -
zeromodels/mobilenetv3_large_100_ra4_e3600_r224_in1k
Image Classification • Updated • 41 -
zeromodels/mobilenetv3_large_100_ra_in1k
Image Classification • Updated • 31
Pure-Keras 3 Next-ViT image-classification weights for kerasformers.
Pure-Keras 3 PiT image-classification weights for kerasformers.
Pure-Keras 3 PoolFormer image-classification weights for kerasformers.
Pure-Keras 3 Res2Net image-classification weights for kerasformers.
Pure-Keras 3 ResMLP image-classification weights for kerasformers.
Pure-Keras 3 ResNet image-classification weights for kerasformers.
Pure-Keras 3 ResNetV2 (BiT) image-classification weights for kerasformers.
-
zeromodels/resnetv2_101x1_bit_goog_in21k
Image Classification • Updated • 38 -
zeromodels/resnetv2_101x1_bit_goog_in21k_ft_in1k
Image Classification • Updated • 44 -
zeromodels/resnetv2_101x3_bit_goog_in21k
Image Classification • Updated • 580 -
zeromodels/resnetv2_101x3_bit_goog_in21k_ft_in1k
Image Classification • Updated • 44
Pure-Keras 3 ResNeXt image-classification weights for kerasformers.
-
zeromodels/resnext101_32x16d_fb_ssl_yfcc100m_ft_in1k
Image Classification • Updated • 54 -
zeromodels/resnext101_32x16d_fb_swsl_ig1b_ft_in1k
Image Classification • Updated • 39 -
zeromodels/resnext101_32x16d_fb_wsl_ig1b_ft_in1k
Image Classification • Updated • 49 -
zeromodels/resnext101_32x32d_fb_wsl_ig1b_ft_in1k
Image Classification • Updated • 49
Pure-Keras 3 SE-ResNet and SE-ResNeXt image-classification weights for kerasformers.
Pure-Keras 3 Swin Transformer image-classification weights for kerasformers.
-
zeromodels/swin_base_patch4_window12_384_ms_in1k
Image Classification • Updated • 44 -
zeromodels/swin_base_patch4_window12_384_ms_in22k
Image Classification • Updated • 49 -
zeromodels/swin_base_patch4_window12_384_ms_in22k_ft_in1k
Image Classification • Updated • 51 -
zeromodels/swin_base_patch4_window7_224_ms_in1k
Image Classification • Updated • 42
Pure-Keras 3 Swin Transformer V2 image-classification weights for kerasformers.
-
zeromodels/swinv2_base_window12_192_ms_in22k
Image Classification • Updated • 44 -
zeromodels/swinv2_base_window12to16_192to256_ms_in22k_ft_in1k
Image Classification • Updated • 41 -
zeromodels/swinv2_base_window12to24_192to384_ms_in22k_ft_in1k
Image Classification • Updated • 33 -
zeromodels/swinv2_base_window16_256_ms_in1k
Image Classification • Updated • 30
Pure-Keras 3 EfficientNet image-classification weights for kerasformers.
Pure-Keras 3 EfficientNet-Lite image-classification weights for kerasformers.
Pure-Keras 3 EfficientNetV2 image-classification weights for kerasformers.
Pure-Keras 3 VGG image-classification weights for kerasformers.
Pure-Keras 3 ViT image-classification weights for kerasformers.
-
zeromodels/vit_base_patch16_224_augreg_in1k
Image Classification • Updated • 39 -
zeromodels/vit_base_patch16_224_augreg_in21k
Image Classification • Updated • 46 -
zeromodels/vit_base_patch16_224_augreg_in21k_ft_in1k
Image Classification • Updated • 52 -
zeromodels/vit_base_patch16_224_orig_in21k_ft_in1k
Image Classification • Updated • 43
Pure-Keras 3 Xception image-classification weights for kerasformers.
Pure-Keras 3 MiT (SegFormer encoder) weights (classification and segmentation) for kerasformers.
Pure-Keras 3 MobileViT weights (classification and segmentation) for kerasformers.
Pure-Keras 3 MobileViT-V2 weights (classification and segmentation) for kerasformers.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of RoBERTa encoders and masked-LM heads.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of BERT encoders and masked-LM heads.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the multilingual XLM-RoBERTa encoders and masked-LM heads.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of DeBERTa, DeBERTa-v2, and DeBERTa-v3 encoders.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the DeepSeek-VL vision-language models.
Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the DeepSeek Janus-Pro any-to-any multimodal models.
Pure-Keras 3 GPT and GPT-2 language models for kerasformers.
OpenAI's open-weight GPT-OSS mixture-of-experts LLMs converted to pure Keras 3 for kerasformers, experts kept in MXFP4.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Google's Gemma 3, for kerasformers.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Google's Gemma 4, for kerasformers.
kerasformers ports of Google's Gemma 3n on-device multimodal models (image + audio + text), runnable on JAX / PyTorch / TensorFlow.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Qwen 2.5 VL for kerasformers.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Qwen 2 VL for kerasformers.
Pure-Keras 3 conversions of Qwen3-Next-80B-A3B (kerasformers).
Pure-Keras 3 conversions of Qwen3.5-MoE (kerasformers).
Pure-Keras 3 conversions of Qwen3-VL-MoE (kerasformers).
Pure-Keras 3 conversions of Qwen2-MoE sparse Mixture-of-Experts LLMs (incl. Qwen1.5-MoE) (kerasformers).
Pure-Keras 3 conversions of Qwen3-MoE sparse Mixture-of-Experts LLMs (kerasformers).
Pure-Keras 3 conversions of Qwen3.8 (kerasformers).
Pure-Keras 3 conversions of InternVL 3 and 3.5 (kerasformers).
Pure-Keras 3 conversions of GLM (kerasformers).
Pure-Keras 3 conversions of the original T5 (kerasformers).
Google's ELECTRA text encoders in pure Keras 3 (kerasformers), runnable on JAX / PyTorch / TensorFlow.
Pure-Keras 3 (JAX / PyTorch / TensorFlow) MobileNetV4 checkpoints for kerasformers, converted from timm.
-
zeromodels/mobilenetv4_conv_small_e2400_r224_in1k
Image Classification • Updated • 38 -
zeromodels/mobilenetv4_conv_medium_e500_r256_in1k
Image Classification • Updated • 39 -
zeromodels/mobilenetv4_conv_large_e600_r384_in1k
Image Classification • Updated • 29 -
zeromodels/mobilenetv4_hybrid_medium_e500_r224_in1k
Image Classification • Updated • 36
Pure-Keras 3 (JAX/PyTorch/TF) TIPSv2 vision-language dual encoders, converted from google/tipsv2
Pure-Keras 3 (JAX/PyTorch/TF) TIPSv2-DPT depth + segmentation models converted from google/tipsv2-dpt