Hugging Face's logo

goldhub
/
Qwen3.8-27B-INT4-W4A16-AutoRound

Image-Text-to-Text
Safetensors
qwen3_5
27b
qwen
qwen3.8
qwen3.8-27b
vision
video-text-to-text
quantized
auto-round
w4a16
mtp
256k-context
conversational
4-bit precision
Model card Files
xet
 Community
1
New discussion

We Cracked Qwen3.8-27B Quant: 27GB INT4 that actually thinks (Goldhub Edition)

13
#1 opened 3 days ago by
goldhub