HK
Heykuki News
Top
New
Best
Ask
Show
Jobs
Toggle theme
Top
New
Best
Ask
Show
Jobs
Request
1.
▲
Ask HN: How to Discover Deep Learning Research Labs?
discuss
5 years ago
bluu00
1 points
2.
▲
How attention sinks keep language models stable
hanlab.mit.edu
36 comments
a year ago
pr337h4m
219 points
3.
▲
SVDQuant: 4-Bit Quantization Powers 12B Flux on a 16GB 4090 GPU with 3x Speedup
hanlab.mit.edu
65 comments
2 years ago
lmxyy
179 points
4.
▲
SVDQuant+NVFP4: 4× Smaller, 3× Faster FLUX with 16-bit Quality on Blackwell GPUs
hanlab.mit.edu
10 comments
a year ago
lmxyy
52 points
5.
▲
Radial Attention: O(nlogn) Attention for Long Video Generation with 2-4× Speedup
hanlab.mit.edu
1 comment
a year ago
lmxyy
6 points
6.
▲
RTX 5090 Workstation Configuration Journey
hanlab.mit.edu
1 comment
a year ago
lmxyy
5 points
7.
▲
TinyChat: Large Language Model on the Edge
hanlab.mit.edu
1 comment
3 years ago
enduku
2 points
8.
▲
HART: Efficient Visual Generation with Hybrid Autoregressive Transformer
hanlab.mit.edu
discuss
2 years ago
lnyan
2 points