debug yolov3

run_yolov3_8card.py

@@ -0,0 +1,17 @@
+import os
+
+os.system(""" ls /home && echo "show code" && ls /tmp/code && echo "show dataset" && ls /tmp/dataset && echo "show output" && ls /tmp/output  && \
+    ln -sf /tmp/dataset ./TopsRider_t2x_2.1.52_samples/samples/dataset && echo "link dataset done" && \
+    ls ./TopsRider_t2x_2.1.52_samples/samples/dataset/ && \
+    bash ./TopsRider_t2x_2.1.52_samples/samples/model/torch/distributed/run_distributed_pytorch_ddp_yolov3_coco2017_convergence_ef32_test.sh && \
+    echo "train done" 
+    """)  
+
+os.system("""
+    cp -r ./TopsRider_t2x_2.1.52_samples/samples/model/torch/distributed/logs /tmp/output && \
+    efsmi -q | grep -E "DEV ID|SN" && \
+    efsmi -dic | tee /tmp/output/dic.log && \
+    cp -r ./TopsRider_t2x_2.1.52_samples/models/training/official/pytorch/vision/detection/yolov3/runs /tmp/output && \
+    ls /tmp/output && grep "train_fps_mean" /tmp/output/logs/*.log && \
+    tail -100 /tmp/output/logs/*.log
+    """)