# --- Rust ---
/target/
**/target/

# --- Python / wheels ---
__pycache__/
*.py[cod]
*.pyd
*.so
*.dylib
*.dll
.venv/
venv/
.mypy_cache/
.pytest_cache/
.ruff_cache/

# maturin / packaging outputs
/dist/
**/*.egg-info/

# --- Large data / generated artifacts (do not commit) ---
*.log
/length_tokenizer.log
*.png
*.pdf
*.jpg
*.webp
*.csv
*.out

# corpora
/corpus_*.txt
!/corpus_small.txt
!/corpus_smoke.txt

# derived inspection outputs
/corpus_inspect_out/
/corpus_py_initial_chars*.tsv

# token tables (can be regenerated; some are multi-GB)
/token_table*.json
/token_table_safe*.json

## training / experiment outputs (large; reproducible)
/run_llama_*/
/run_llama_*.log
/pipeline_*.log
/driver_*.log
/loss_curve*
/loss_overlay*
/bpc_overlay*
/loss_sync_*/
/ckpt_*/
/model_*/
/tokenizer_out_*/
/exp_*/
/dist_default_*/
/dist_full_*/
/_ckpt_test_save/




