DeBERTa (v1 / v2 / v3) Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of DeBERTa, DeBERTa-v2, and DeBERTa-v3 encoders. Collection by zeromodels Aug 26 - zeromodels/deberta_base Fill-Mask • Updated Aug 26 • 19 zeromodels/deberta_large Fill-Mask • Updated Aug 26 • 12 zeromodels/deberta_v2_xlarge Fill-Mask • Updated Aug 26 • 9 zeromodels/deberta_v2_xxlarge Fill-Mask • Updated Aug 26 • 5
Janus-Pro Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the DeepSeek Janus-Pro any-to-any multimodal models. Collection by zeromodels Aug 26 - zeromodels/janus_pro_1b Image-Text-to-Text • Updated Sep 29 • 50 zeromodels/janus_pro_7b Image-Text-to-Text • Updated Aug 26 • 2
GPT-OSS OpenAI's open-weight GPT-OSS mixture-of-experts LLMs converted to pure Keras 3 for kerasformers, experts kept in MXFP4. Collection by zeromodels Aug 26 - zeromodels/gpt-oss-20b Text Generation • Updated Aug 26 • 4 zeromodels/gpt-oss-120b Text Generation • Updated Aug 26 • 6
Gemma 4 Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Google's Gemma 4, for kerasformers. Collection by zeromodels Aug 26 - zeromodels/gemma-4-e4b-it Any-to-Any • Updated Aug 26 • 7 zeromodels/gemma-4-e2b-it Any-to-Any • Updated Aug 26 • 6 zeromodels/gemma-4-31b-it Image-Text-to-Text • Updated Aug 26 • 2 zeromodels/gemma-4-26b-a4b-it Image-Text-to-Text • Updated Aug 26 • 3
Qwen2.5-VL Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Qwen 2.5 VL for kerasformers. Collection by zeromodels Aug 26 - zeromodels/qwen2.5-vl-3b-instruct Image-Text-to-Text • Updated Aug 26 • 6 zeromodels/qwen2.5-vl-7b-instruct Image-Text-to-Text • Updated Aug 26 • 8 zeromodels/qwen2.5-vl-32b-instruct Image-Text-to-Text • Updated Aug 26 • 6 zeromodels/qwen2.5-vl-72b-instruct Image-Text-to-Text • Updated Aug 26 • 5
DeepSeek-VL Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the DeepSeek-VL vision-language models. Collection by zeromodels Aug 26 - zeromodels/deepseek_vl_1.3b_chat Image-Text-to-Text • Updated Sep 29 • 31 zeromodels/deepseek_vl_1.3b_base Image-Text-to-Text • Updated Sep 29 • 16 zeromodels/deepseek_vl_7b_chat Image-Text-to-Text • Updated Aug 26 • 6 zeromodels/deepseek_vl_7b_base Image-Text-to-Text • Updated Aug 26 • 10
GPT and GPT-2 Pure-Keras 3 GPT and GPT-2 language models for kerasformers. Collection by zeromodels Aug 26 - zeromodels/gpt Text Generation • Updated Aug 26 • 9 zeromodels/gpt2 Text Generation • Updated Aug 26 • 9 zeromodels/gpt2_medium Text Generation • Updated Aug 26 • 9 zeromodels/gpt2_large Text Generation • Updated Sep 29 • 4
Gemma 3 Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Google's Gemma 3, for kerasformers. Collection by zeromodels Aug 26 - zeromodels/gemma-3-1b-it Text Generation • Updated Aug 26 • 21 • 1 zeromodels/gemma-3-1b-pt Text Generation • Updated Aug 26 • 29 zeromodels/gemma-3-4b-it Image-Text-to-Text • Updated Aug 26 • 2 zeromodels/gemma-3-4b-pt Image-Text-to-Text • Updated Aug 26 • 5
Gemma 3n kerasformers ports of Google's Gemma 3n on-device multimodal models (image + audio + text), runnable on JAX / PyTorch / TensorFlow. Collection by zeromodels Aug 26 - zeromodels/gemma-3n-e2b Image-Text-to-Text • Updated Sep 29 • 11 zeromodels/gemma-3n-e2b-it Image-Text-to-Text • Updated Sep 29 • 11 zeromodels/gemma-3n-e4b Image-Text-to-Text • Updated Sep 29 • 6 zeromodels/gemma-3n-e4b-it Image-Text-to-Text • Updated Sep 29 • 9
Qwen2-VL Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Qwen 2 VL for kerasformers. Collection by zeromodels Aug 26 - zeromodels/qwen2-vl-2b Image-Text-to-Text • Updated Aug 26 • 24 zeromodels/qwen2-vl-2b-instruct Image-Text-to-Text • Updated Aug 26 • 43 zeromodels/qwen2-vl-7b-instruct Image-Text-to-Text • Updated Aug 26 • 9 zeromodels/qwen2-vl-7b Image-Text-to-Text • Updated Aug 26 • 3
DeBERTa (v1 / v2 / v3) Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of DeBERTa, DeBERTa-v2, and DeBERTa-v3 encoders. Collection by zeromodels Aug 26 - zeromodels/deberta_base Fill-Mask • Updated Aug 26 • 19 zeromodels/deberta_large Fill-Mask • Updated Aug 26 • 12 zeromodels/deberta_v2_xlarge Fill-Mask • Updated Aug 26 • 9 zeromodels/deberta_v2_xxlarge Fill-Mask • Updated Aug 26 • 5
DeepSeek-VL Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the DeepSeek-VL vision-language models. Collection by zeromodels Aug 26 - zeromodels/deepseek_vl_1.3b_chat Image-Text-to-Text • Updated Sep 29 • 31 zeromodels/deepseek_vl_1.3b_base Image-Text-to-Text • Updated Sep 29 • 16 zeromodels/deepseek_vl_7b_chat Image-Text-to-Text • Updated Aug 26 • 6 zeromodels/deepseek_vl_7b_base Image-Text-to-Text • Updated Aug 26 • 10
Janus-Pro Pure Keras 3, backend-agnostic (JAX / TensorFlow / PyTorch) ports of the DeepSeek Janus-Pro any-to-any multimodal models. Collection by zeromodels Aug 26 - zeromodels/janus_pro_1b Image-Text-to-Text • Updated Sep 29 • 50 zeromodels/janus_pro_7b Image-Text-to-Text • Updated Aug 26 • 2
GPT and GPT-2 Pure-Keras 3 GPT and GPT-2 language models for kerasformers. Collection by zeromodels Aug 26 - zeromodels/gpt Text Generation • Updated Aug 26 • 9 zeromodels/gpt2 Text Generation • Updated Aug 26 • 9 zeromodels/gpt2_medium Text Generation • Updated Aug 26 • 9 zeromodels/gpt2_large Text Generation • Updated Sep 29 • 4
GPT-OSS OpenAI's open-weight GPT-OSS mixture-of-experts LLMs converted to pure Keras 3 for kerasformers, experts kept in MXFP4. Collection by zeromodels Aug 26 - zeromodels/gpt-oss-20b Text Generation • Updated Aug 26 • 4 zeromodels/gpt-oss-120b Text Generation • Updated Aug 26 • 6
Gemma 3 Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Google's Gemma 3, for kerasformers. Collection by zeromodels Aug 26 - zeromodels/gemma-3-1b-it Text Generation • Updated Aug 26 • 21 • 1 zeromodels/gemma-3-1b-pt Text Generation • Updated Aug 26 • 29 zeromodels/gemma-3-4b-it Image-Text-to-Text • Updated Aug 26 • 2 zeromodels/gemma-3-4b-pt Image-Text-to-Text • Updated Aug 26 • 5
Gemma 4 Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Google's Gemma 4, for kerasformers. Collection by zeromodels Aug 26 - zeromodels/gemma-4-e4b-it Any-to-Any • Updated Aug 26 • 7 zeromodels/gemma-4-e2b-it Any-to-Any • Updated Aug 26 • 6 zeromodels/gemma-4-31b-it Image-Text-to-Text • Updated Aug 26 • 2 zeromodels/gemma-4-26b-a4b-it Image-Text-to-Text • Updated Aug 26 • 3
Gemma 3n kerasformers ports of Google's Gemma 3n on-device multimodal models (image + audio + text), runnable on JAX / PyTorch / TensorFlow. Collection by zeromodels Aug 26 - zeromodels/gemma-3n-e2b Image-Text-to-Text • Updated Sep 29 • 11 zeromodels/gemma-3n-e2b-it Image-Text-to-Text • Updated Sep 29 • 11 zeromodels/gemma-3n-e4b Image-Text-to-Text • Updated Sep 29 • 6 zeromodels/gemma-3n-e4b-it Image-Text-to-Text • Updated Sep 29 • 9
Qwen2.5-VL Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Qwen 2.5 VL for kerasformers. Collection by zeromodels Aug 26 - zeromodels/qwen2.5-vl-3b-instruct Image-Text-to-Text • Updated Aug 26 • 6 zeromodels/qwen2.5-vl-7b-instruct Image-Text-to-Text • Updated Aug 26 • 8 zeromodels/qwen2.5-vl-32b-instruct Image-Text-to-Text • Updated Aug 26 • 6 zeromodels/qwen2.5-vl-72b-instruct Image-Text-to-Text • Updated Aug 26 • 5
Qwen2-VL Pure-Keras 3 (JAX / PyTorch / TensorFlow) conversions of Qwen 2 VL for kerasformers. Collection by zeromodels Aug 26 - zeromodels/qwen2-vl-2b Image-Text-to-Text • Updated Aug 26 • 24 zeromodels/qwen2-vl-2b-instruct Image-Text-to-Text • Updated Aug 26 • 43 zeromodels/qwen2-vl-7b-instruct Image-Text-to-Text • Updated Aug 26 • 9 zeromodels/qwen2-vl-7b Image-Text-to-Text • Updated Aug 26 • 3
Aug 26