-
Notifications
You must be signed in to change notification settings - Fork 2
/
requirements.txt
61 lines (41 loc) · 1.4 KB
/
requirements.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
# requirements.txt for BitNet CUDA project
# CUDA toolkit (not pip-installable, must be installed separately)
# Minimum version: CUDA 11.0
# PyTorch with CUDA support
torch>=1.7.0
# NumPy for CPU-based operations
numpy>=1.19.0
# NVIDIA APEX for mixed precision training (optional)
apex
# CuPy for additional CUDA operations
cupy-cuda110 # Use appropriate version matching your CUDA installation
# Pytest for running tests
pytest>=6.0.0
# Matplotlib for visualization (if needed)
matplotlib>=3.3.0
# tqdm for progress bars
tqdm>=4.50.0
# Pillow for image processing (if needed)
Pillow>=8.0.0
# PyYAML for configuration file parsing
PyYAML>=5.3.0
# Tensorboard for logging and visualization
tensorboard>=2.4.0
# NVIDIA DALI for data loading and preprocessing (optional)
nvidia-dali-cuda110 # Use appropriate version matching your CUDA installation
# Numba for JIT compilation (optional)
numba>=0.51.0
# Pandas for data manipulation (if needed)
pandas>=1.1.0
# SciPy for scientific computing (if needed)
scipy>=1.5.0
# Jupyter for interactive development (optional)
jupyter
# Black for code formatting (development only)
black
# Flake8 for linting (development only)
flake8
# MyPy for static type checking (development only)
mypy
# NVIDIA Nsight Systems for profiling (not pip-installable, must be installed separately)
# NVIDIA Nsight Compute for kernel analysis (not pip-installable, must be installed separately)