fix dlrm when using incbench (#2084)

Signed-off-by: Xin He <xinhe3@habana.ai>
intel · Dec 6, 2024 · 9118449 · 9118449
1 parent 619f77b
commit 9118449
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/examples/3.x_api/pytorch/recommendation/dlrm/static_quant/ipex/run_benchmark.sh b/examples/3.x_api/pytorch/recommendation/dlrm/static_quant/ipex/run_benchmark.sh
@@ -80,7 +80,7 @@ function run_tuning {
         --save-model ${tuned_checkpoint} --test-freq=2048 --print-auc $ARGS \
         --load-model=${input_model}  --accuracy_only
     elif [[ ${mode} == "performance" ]]; then
-        incbench --num_cores_per_instance 4 -u $MODEL_SCRIPT \
+        incbench --num_cores_per_instance 4 $MODEL_SCRIPT \
         --raw-data-file=${dataset_location}/day --processed-data-file=${dataset_location}/terabyte_processed.npz \
         --data-set=terabyte --benchmark \
         --memory-map --mlperf-bin-loader --round-targets=True --learning-rate=1.0 \