Local Logging & Tracking¶
This repository uses local-only logging:
MLflow (file backend)¶
- Default:
MLFLOW_TRACKING_URIis set tofile:./mlrunswhen using helpers inlogging_utils.py. - To customize:
References: MLflow file backend / tracking URIs. (mlflow.org)
TensorBoard¶
Use PyTorch’s SummaryWriter:
from torch.utils.tensorboard import SummaryWriter
writer = SummaryWriter("./runs")
writer.add_scalar("loss", 0.123, 1)
Reference: PyTorch TensorBoard utilities. (docs.pytorch.org)
NDJSON & CSV outputs¶
- NDJSON: one JSON object per line; prefer the
.ndjsonextension. (GitHub) - The evaluation logger now reuses the training NDJSON sink, supporting file
rotation (
ndjson_max_bytes,ndjson_backup_count) and optional GPU telemetry whenpynvmlis installed. - CSV: use Python’s
csv.DictWriterfor tabular outputs. (docs.python.org)