feat(streaming): CONDA_ENV knob for the matrix submitter

Add CONDA_ENV=<name> to run each matrix job via `conda run --no-capture-output -n
<name>` — works inside the dash `sbatch --wrap` without sourcing conda.sh / activating,
and streams logs live. Point it at a conda env with a modern torchcodec (>=0.11) +
datasets (>=4.7); the default cluster `base` env is often too old to decode AV1.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
Pepijn
2026-06-09 16:55:42 +02:00
parent 23c58f5f9e
commit 06ddc59913
+7
View File
@@ -38,6 +38,13 @@ CPU_BUFFER=${CPU_BUFFER:-4000}
GPU_BUFFER=${GPU_BUFFER:-1000} # smaller buffer bounds on-GPU frame memory
BATCH_SIZE=${BATCH_SIZE:-64}
RUN=${RUN:-python}
# CONDA_ENV=<name> runs each job via `conda run -n <name>` (no activation needed inside the dash --wrap;
# --no-capture-output streams logs live). Set this to a conda env that has a MODERN torchcodec (>=0.11)
# + datasets (>=4.7) — the default `base` env on many clusters is too old to decode AV1 / lacks CUDA.
CONDA_ENV=${CONDA_ENV:-}
if [ -n "$CONDA_ENV" ] && [ "$RUN" = "python" ]; then
RUN="conda run --no-capture-output -n $CONDA_ENV python"
fi
SOURCES=${SOURCES:-"hub bucket warmed_bucket"}
MODES=${MODES:-"single sarm"}