issues
search
MrYxJ
/
calculate-flops.pytorch
The calflops is designed to calculate FLOPs、MACs and Parameters in all various neural networks, such as Linear、 CNN、 RNN、 GCN、Transformer(Bert、LlaMA etc Large Language Model)
https://pypi.org/project/calflops/
MIT License
583
stars
21
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Compute the FLOPs for each transformer layer of an LLM
#43
jeremyzhangsq
opened
3 weeks ago
0
Ran the example code but got wrong results
#42
nameongithub
opened
1 month ago
2
fix bug in _einsum_flops_compute when using torch.einsum
#41
Necolizer
opened
1 month ago
0
AttributeError: 'list' object has no attribute 'shape' in function _einsum_flops_compute()
#40
Necolizer
opened
1 month ago
1
How to test models deployed on multiple GPUs？
#39
May2064
opened
2 months ago
0
packaging.version.InvalidVersion: Invalid version: '0.10.1,<0.11'
#38
ymmm-4
opened
2 months ago
0
Flops for GNN
#37
ritamartinspereira
opened
2 months ago
1
Why are `torch.mm` and `torch.bmm` deleted in pytorch_ops.py
#36
wzk1015
opened
3 months ago
2
Only nn.modules are supported for FLOP analysis?
#35
avisinghal6
opened
4 months ago
0
Add support of non-linear operations
#34
Ther-nullptr
opened
4 months ago
0
MACs for ConvTranspose2d
#33
qiuzidian
opened
4 months ago
0
Formatting and docstring fixes
#32
bhavika
closed
5 months ago
0
gated models do not work
#31
mohbay
opened
5 months ago
1
How to calculate the FLOPs when the model has some freezed variables, or using LoRA?
#30
Ledzy
opened
5 months ago
0
On the calculation of FLOPs for SAM/MobileSAM
#29
BeliverK
opened
5 months ago
0
评价Gemma的时候参数量为0
#28
JackeyWang777
closed
5 months ago
1
关于百分比的问题
#27
lzd-1230
closed
5 months ago
2
你好，比如llama这些LLM模型，我看统计的结果好像没有包括attention矩阵的Q@K.T那部分计算量。
#26
ywh187
opened
5 months ago
5
Is it considering output length of transformer models?
#25
liut89
closed
5 months ago
1
flops percent wrong
#24
Lix1993
closed
5 months ago
1
Can this support StableDiffusion？
#23
sunzhe09
opened
6 months ago
0
[Bug] Upsample flops not being assigned after calculation
#22
Ridhamz-nd
opened
6 months ago
0
Add sentence-transformers
#21
bhavika
closed
6 months ago
0
What is the difference between the 'forward' mode and the 'generate' mode?
#20
AboveParadise
opened
7 months ago
1
Please make a CITATION file
#19
Livioni
closed
5 months ago
1
Library `gpt-neox` is not supported yet
#18
melisa-writer
closed
5 months ago
1
How to calculate model with more than one input?
#17
yuanzhi-zhu
opened
8 months ago
0
Why the unit of fwd FLOPs is FLOPS?
#16
Fibonacci43
opened
9 months ago
1
allow for non-tensor kwargs in calculate_flops
#15
vzantedeschi
closed
9 months ago
1
error with the example code
#14
tranlm
closed
9 months ago
1
about 3D object detection
#13
ouyangziyao
opened
12 months ago
0
Library `diffusers` is not supported yet
#12
XianweiC
opened
1 year ago
2
Cannot calculate EfficientNet
#11
QingyuLiu
closed
5 months ago
1
访存量的计算
#10
Amanda-Barbara
closed
10 months ago
2
why I can not find transformers.modeling_outputs,Can you tell me your vision of transformers
#9
kingofengineer
closed
1 year ago
1
why seq_len of llm is 128?
#8
LinglingGreat
closed
1 year ago
1
Why is the proportion of FLOPs in each submodule always half of the proportion of MACS
#7
wulinunu
opened
1 year ago
1
ImportError: cannot import name 'is_timm_available' from 'accelerate.utils'
#6
jiabao-wang
closed
1 year ago
2
This project can be used on large visual language models?
#5
jiabao-wang
closed
1 year ago
4
Support sparse model calculation
#4
pingzhili
closed
1 year ago
3
给的示例无法运行
#3
Arcmoon-Hu
closed
1 year ago
1
ImportError at 'from calflops import calculate_flops': cannot import name 'is_timm_available' from 'accelerate.utils'
#2
biplob1ly
closed
1 year ago
3
大模型微调
#1
Pobby321
closed
1 year ago
1