A pytorch perf decorator
Project description
PyTorch Perf
Install
pip install -U pytorch-perf
To verify a successful installation:
python -c "from torchperf import info; info.show()"
Usage
import torch
from torchperf import perf, info
info.show()
N = 100
x = torch.rand(N, N, device="cuda")
results = []
repeats = 100
@perf(o=results, n=repeats)
def mul(x, y):
return x * y
z = mul(x, x)
print(results[:5])
print("avg:", sum(results)/len(results))
@perf(n=repeats)
def mul(x, y):
return x * y
z = mul(x, x)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
pytorch-perf-0.0.2.tar.gz
(5.6 kB
view details)
Built Distribution
File details
Details for the file pytorch-perf-0.0.2.tar.gz
.
File metadata
- Download URL: pytorch-perf-0.0.2.tar.gz
- Upload date:
- Size: 5.6 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.9.13
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 0655d13ab913ea186e69a36a0d62cdd9cc6a1e399ea51af851a2452b133dae7b |
|
MD5 | 3bab61e57c5bb7b8cb609eada2d030fd |
|
BLAKE2b-256 | ac27b8c14f755874af3cba8c8dce81c995f9fad234a8ad52558ac42f72ac5e9a |
File details
Details for the file pytorch_perf-0.0.2-py3-none-any.whl
.
File metadata
- Download URL: pytorch_perf-0.0.2-py3-none-any.whl
- Upload date:
- Size: 5.4 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/4.0.2 CPython/3.9.13
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 30935b4767f358c8b94c2dab7836c1a08c542a90e5cc18c4b1874c5c2e28649c |
|
MD5 | 926160931937d2820e1bffe7ef03e672 |
|
BLAKE2b-256 | 129172a586ad25774ba8d4c31313e5cf30e3a23d3f8fb91752a82f9a06b52335 |