Add use_gpu for PureSVD

[chezou]

Description

Introduce use_gpu for PureSVD.

My microbenchmark code shows 8 times faster than CPU with Movielens 1M dataset on GeForce 3060Ti compared with Ryzen 5 5600X.

>>> import timeit
>>> with open("scripts/benchmark.py", "r") as f:
...     code = f.read()
...
>>> print(code)
import pandas as pd

from rectools import Columns
from rectools.dataset import Dataset
from rectools.models.pure_svd import PureSVDModel

ratings = pd.read_csv(
    "ml-1m/ratings.dat",
    sep="::",
    engine="python",  # Because of 2-chars separators
    header=None,
    names=[Columns.User, Columns.Item, Columns.Weight, Columns.Datetime],
)

# Create dataset
dataset = Dataset.construct(ratings)

model = PureSVDModel(factors=100, use_gpu={})
>>> timeit.timeit("model.fit(dataset)", setup=code.format(False), number=10)
42.881751499007805
>>> timeit.timeit("model.fit(dataset)", setup=code.format(True), number=10)
5.059483032004209

Environment information:

➜  RecTools git:(svd-gpu) ✗ poetry run python -c  "import cupy; cupy.show_config(_full=True)"
OS                           : Linux-5.15.167.4-microsoft-standard-WSL2-x86_64-with-glibc2.39
Python Version               : 3.12.3
CuPy Version                 : 13.3.0
CuPy Platform                : NVIDIA CUDA
NumPy Version                : 1.26.4
SciPy Version                : 1.12.0
Cython Build Version         : 0.29.36
Cython Runtime Version       : None
CUDA Root                    : /usr/local/cuda
nvcc PATH                    : /usr/local/cuda/bin/nvcc
CUDA Build Version           : 12060
CUDA Driver Version          : 12060
CUDA Runtime Version         : 12060 (linked to CuPy) / 12060 (locally installed)
CUDA Extra Include Dirs      : []
cuBLAS Version               : 120604
cuFFT Version                : 11300
cuRAND Version               : 10307
cuSOLVER Version             : (11, 7, 1)
cuSPARSE Version             : 12504
NVRTC Version                : (12, 6)
Thrust Version               : 200600
CUB Build Version            : 200600
Jitify Build Version         : <unknown>
cuDNN Build Version          : None
cuDNN Version                : None
NCCL Build Version           : None
NCCL Runtime Version         : None
cuTENSOR Version             : None
cuSPARSELt Build Version     : None
Device 0 Name                : NVIDIA GeForce RTX 3060 Ti
Device 0 Compute Capability  : 86
Device 0 PCI Bus ID          : 0000:09:00.0

Type of change

• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Optimization

How Has This Been Tested?

Before submitting a PR, please check yourself against the following list. It would save us quite a lot of time.

• Have you read the contribution guide?
• Have you updated the relevant docstrings? We're using Numpy format, please double-check yourself
• Does your change require any new tests?
• Have you updated the changelog file?

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 100.00%. Comparing base (9b3992e) to head (d19a277).
Report is 91 commits behind head on main.

Additional details and impacted files

@@             Coverage Diff             @@
##              main      #229     +/-   ##
===========================================
  Coverage   100.00%   100.00%             
===========================================
  Files           45        60     +15     
  Lines         2242      4016   +1774     
===========================================
+ Hits          2242      4016   +1774     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[feldlime]

Thanks for the PR, it looks super cool!

[chezou]

Python 3.9 CI fails because of codecov issue. Test is passed.

error - 2025-01-16 17:47:40,999 -- Report creating failed: {"message":"Token required because branch is protected"}

[chezou]

@feldlime I addressed all the comments you made. PTAL