Joon
mconcat
AI & ML interests
Interpretability, RL
Organizations
Avg Draft acceptance rate: 0.0%
2
#1 opened 2 months ago
by
BuiDoan
Any way to mtp?
4
#1 opened 4 months ago
by
tasticleeze
MTP doesn't work
2
#1 opened 4 months ago
by
Serpov
AWQ 4/5/6-bit request for Qwopus3.5-27B-v3
1
#2 opened 4 months ago
by
celikburak
Custom tokenizer
1
#1 opened 4 months ago
by
BuiDoan
AWQ 4-bit version of this Opus-Distilled-v2 model?
4
#1 opened 4 months ago
by
celikburak
Blackwell (RTX 5090 and RTX 6000) compatibility?
🤝 1
11
#1 opened 5 months ago
by
jhsmith0
MTP and calibration dataset
2
#4 opened 5 months ago
by
jmckenzie-dev
Can you release an FP8 version for vLLM as well please? NVFP4 might not be so good.
4
#2 opened 5 months ago
by
AQLabs
Blackwell (RTX 5090 and RTX 6000) compatibility?
🤝 1
11
#1 opened 5 months ago
by
jhsmith0
Can you release an FP8 version for vLLM as well please? NVFP4 might not be so good.
4
#2 opened 5 months ago
by
AQLabs
Blackwell (RTX 5090 and RTX 6000) compatibility?
🤝 1
11
#1 opened 5 months ago
by
jhsmith0