-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathrequirements.txt
More file actions
74 lines (59 loc) · 2.17 KB
/
Copy pathrequirements.txt
File metadata and controls
74 lines (59 loc) · 2.17 KB
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
62
63
64
65
66
67
68
69
70
71
72
73
74
# =============================================================================
# Full Development Environment Dependencies
# =============================================================================
# This file lists dependencies for a complete local dev/CI environment.
# It is NOT the package's published dependency list; that lives in
# [project.dependencies] inside pyproject.toml (minimal set).
# Install with: pip install -r requirements.txt
# =============================================================================
# Build toolchain
cmake>=3.24
cython>=0.28
# Hardware modeling (keep in sync with pyproject.toml [project.dependencies])
accelforge>=1.0.355
# Formal verification (keep in sync with pyproject.toml [project.dependencies])
z3-solver>=4.12.0
torch>=2.4
accelerate==1.8.0
numpy
graphviz
tqdm
transformers==4.53.0
protobuf==3.20.0
tg4perfetto==0.0.4
# Distributed computing (Ray) — part of core dependencies in pyproject.toml
ray>=2.55.0
# =============================================================================
# Backend-Specific Dependencies (install as needed)
# =============================================================================
# TPU Backend (Google TPU via JAX)
# pip install jax[tpu] -f https://storage.googleapis.com/jax-releases/libtpu_releases.html
# jax>=0.4.20
# jaxlib>=0.4.20
# flax>=0.8.0
# XPU Backend (Intel GPU via PyTorch)
# pip install intel-extension-for-pytorch
# intel-extension-for-pytorch>=2.1.0
# ROCm Backend (AMD GPU)
# pip install torch-rocm (or build from source with ROCm)
# MACA Backend (MetaX GPU)
# Requires MetaX MACA SDK from vendor
# Ascend Backend (Huawei NPU)
# pip install torch_npu
# torch_npu>=2.1.0
# TVM Backend
# pip install apache-tvm
# apache-tvm>=0.15.0
# MLIR/StableHLO
# pip install mlir-python-bindings (if available)
# Or build MLIR from LLVM source
# Distributed Computing (Ray) — now part of core dependencies (see above)
# ray>=2.9.0
# =============================================================================
# Development Dependencies
# =============================================================================
# pytest>=7.0.0
# pytest-cov>=4.0.0
# black>=23.0.0
# isort>=5.12.0
# mypy>=1.0.0