Gemma 4 QAT models in NVFP4 with FP8-calibrated KV cache using multilingual and tool-use data. 31B, 26B and 12B use imatrix_mse.
Yasuhiro Ohno
yasu-oh
AI & ML interests
None yet
Recent Activity
updated a model about 17 hours ago
yasu-oh/gemma-4-E4B-it-qat-NVFP4 updated a model about 17 hours ago
yasu-oh/gemma-4-E2B-it-qat-NVFP4 updated a collection about 19 hours ago
Gemma 4 QAT NVFP4Organizations
None yet
models 22
yasu-oh/gemma-4-E2B-it-qat-NVFP4
Image-Text-to-Text • Updated
yasu-oh/gemma-4-E4B-it-qat-NVFP4
Image-Text-to-Text • Updated
yasu-oh/gemma-4-12B-it-qat-NVFP4
Image-Text-to-Text • Updated • 202
yasu-oh/gemma-4-26B-A4B-it-qat-NVFP4
Image-Text-to-Text • Updated • 44
yasu-oh/gemma-4-31B-it-qat-NVFP4
Image-Text-to-Text • 18B • Updated • 1.14k • 1
yasu-oh/Qwen3-Reranker-0.6B
Text Ranking • 0.6B • Updated • 6
yasu-oh/pplx-embed-v1-0.6b
Feature Extraction • 0.6B • Updated • 38
yasu-oh/Magistral-Small-2506-GGUF
Text Generation • 24B • Updated • 13
yasu-oh/Llama-3-Swallow-Infused-R1776-70B
Text Generation • 71B • Updated • 6 • 1
yasu-oh/Llama-3-SS-Infused-R1776-70B
Text Generation • 71B • Updated • 6 • 1
datasets 0
None public yet