deokjae
bdbj
AI & ML interests
None yet
Recent Activity
upvoted
a
paper
8 days ago
Efficient Latency-Aware CNN Depth Compression via Two-Stage Dynamic
Programming
updated
a collection
10 days ago
Data-aware quantization w/ Q-Palette
updated
a collection
10 days ago
Data-aware quantization w/ Q-Palette
Organizations
None yet
models
34
bdbj/Llama-3.1-70B-qpal-df-msq-4.25bit
20B
•
Updated
•
11
bdbj/Llama-3.1-70B-qpal-df-msq-4bit
19B
•
Updated
•
9
bdbj/Llama-3.1-70B-qpal-df-msq-3.4bit
16B
•
Updated
•
8
bdbj/Llama-3.1-70B-qpal-df-msq-3.63bit
17B
•
Updated
•
10
bdbj/Llama-3.1-70B-qpal-df-msq-3.25bit
16B
•
Updated
•
11
bdbj/Llama-2-13b-qpal-hess-msq-2bit
2B
•
Updated
•
9
bdbj/Llama-2-13b-qpal-hess-msq-3bit
3B
•
Updated
•
13
bdbj/Llama-2-7b-qpal-hess-msq-3bit
1B
•
Updated
•
15
bdbj/Llama-2-7b-qpal-hess-msq-2bit
1B
•
Updated
•
11
bdbj/Qwen2.5-7B-qpal-df-msq-4.25bit
3B
•
Updated
•
11
datasets
0
None public yet