Local GPU benchmark on 2x RTX 5060 Ti (32GB, vLLM 0.29): 63.4% overall (4 benchmarks) at 151.1 tok/s
#8 opened 24 days ago
by
Mausbiber
Update chat_template.jinja
π 1
1
#7 opened 4 months ago
by
xdavxd
What about an NVFP4 version of the new QAT checkpoints?
π 1
#6 opened 5 months ago
by
j-klesen
TemporalMesh Transformer: 29.4 PPL at 48% compute β beats Mamba, new open-source architecture
#5 opened 6 months ago
by
vigneshwar234
Unable to load using vLLM
2
#3 opened 6 months ago
by
Nithishm2410wsfa
Update chat template
#2 opened 7 months ago
by
allenc87
GGUF available β Cerebellum v3 (11 GB, ablation-guided mixed-precision)
#1 opened 7 months ago
by
deucebucket