DGX_Spark_Unsloth_Lossless_Speedup/.gitignore at main · albond/DGX_Spark_Unsloth_Lossless_Speedup · 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
# Local-only config (paths to private data)
configs/local.env

# Benchmark outputs (keep folders, exclude contents)
results/*
!results/.gitkeep
lab/results/*
!lab/results/.gitkeep
*.log

# Wizard output (merged models — large, user-specific; keep folder, exclude contents)
output/*
!output/.gitkeep

# Don't accidentally commit private data inside the repo
*.jsonl
*.tsv
*.csv

# Python
__pycache__/
*.py[cod]
*.egg-info/
.pytest_cache/
.ruff_cache/
.venv/
venv/

# Editor / OS
.DS_Store
.idea/
.vscode/

# HF / training scratch
training/outputs/
trainer_output/

# Unsloth generates compiled module cache next to train.py on first run.
# These are auto-regenerated, env-specific, and contain torch.compile artifacts —
# they must not be committed.
unsloth_compiled_cache/

# JIT-compiled CUDA extension artifacts (regenerated on first use)
lab/kernels/**/build/
*.so
*.o
*.cu.o

# Local-only working notes / scratch
_archive_internal/
.local-notes/