issues
search
hahnyuan
/
LLM-Viewer
Analyze the inference of Large Language Models (LLMs). Analyze aspects like computation, storage, transmission, and hardware roofline model in a user-friendly interface.
MIT License
275
stars
31
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Tensor parallelism support
#14
UranusSeven
closed
2 days ago
0
-
#13
wutong4012
closed
1 week ago
0
GQA Correction for KV memory
#12
Vashistht
closed
3 weeks ago
1
I am not sure why OPs for softmax is "softmax_OPs = bsz * n_heads * seqlen * 1 * 5"
#11
erxiong0
closed
6 days ago
1
请问hardware_param里面的onchip_buffer指的是哪部分的size？
#10
pangsg
closed
6 days ago
2
Can I import my own model into LLM-Viewer?
#9
Ikonwnothin
opened
2 months ago
0
Error message when running the example command line
#8
amin-rain
opened
3 months ago
1
A40 MAC nubmer
#7
sunshinemyson
closed
4 months ago
1
Support EleutherAI/gpt-j-6B
#6
sunshinemyson
closed
4 months ago
1
[fix] Fix data of qk_matmul in prefill stage
#5
jschenxiaoyu
closed
4 months ago
1
The error between LLM-viewer predicted results and TensorRT-LLM real performance is large.
#4
feifeibear
opened
6 months ago
2
[feat] generation task cli
#3
feifeibear
closed
3 weeks ago
1
How can I get throughput for a generative model
#2
feifeibear
opened
6 months ago
6
[fix] fix analyze_generate_task bug
#1
feifeibear
closed
6 months ago
0