kv-cache-quantization/requirements.txt at main · crack666/kv-cache-quantization · GitHub

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
# Core ML Frameworks
# PyTorch: Installiere mit CUDA 12.1 support via:
# pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
torch>=2.1.0
torchvision>=0.16.0
torchaudio>=2.1.0
transformers==4.35.0
accelerate==0.25.0

# Profiling & Metriken
pynvml==11.5.0
nvidia-ml-py3==7.352.0

# Benchmarks
lm-eval==0.4.0
datasets==2.15.0

# Datenverarbeitung
numpy==1.24.3
pandas==2.1.3
scipy==1.11.4

# Visualisierung
matplotlib==3.8.2
seaborn==0.13.0
plotly==5.18.0

# Experiment-Tracking (optional)
# wandb==0.16.0
# mlflow==2.8.1

# Utilities
tqdm==4.66.1
pyyaml==6.0.1
jsonlines==4.0.0

# Development
ipython==8.18.1
jupyter==1.0.0
black==23.12.0
flake8==6.1.0

# Optional: vLLM (schnelle Inferenz)
# vllm==0.2.5  # Benötigt CUDA 12.x

# Optional: TensorRT-LLM
# tensorrt-llm==0.5.0

# Note: Install PyTorch FIRST with CUDA support:
# pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu121
# Then install rest: pip install -r requirements.txt