YaYaB

Thanks a lot I'll try your fix! I could be a good idea to add unit tests based on expected values for different models predictions to catch those, no?

Actualy it is even not related to tensorrt but even with classical caffe predictions with or without gpu

Honestly, it is kind of pain to install everything in a host as: - it takes time - it adds a lot of unecessary things in my host I would...

@fantes Yes I get a core dump as well with ``` "maxWorkspaceSize" to 2000 ```

I make sure the directory does not contain any of those before creating the serving and launching the prediction. I'll try it in another machine to see If I get...

@fantes Ok my bad. After testing it in an another machine I found that I was using it with a Titan X (a bit older than 1080Ti). When I use...

According to the issue on triton-infenrence-server it should work with cuda 10.1 no?

Using `"engine":"CUDNN_MIN_MEMORY"` ``` curl -X PUT "http://localhost:8080/services/imageserv" -d "{\"mllib\":\"caffe\",\"description\":\"image classification service\",\"type\":\"supervised\",\"parameters\":{\"input\":{\"connector\":\"image\"},\"mllib\":{\"nclasses\":1000, \"engine\":\"CUDNN_MIN_MEMORY\"}},\"model\":{\"repository\":\"/opt/models/ggnet/\"}}" ``` I observe : - 598MiB usage vs 942Mib with cuda 10.1 - 1548MiB usage vs 2804Mib with...

I've built new images without cudnn. Here is what I observe : - 245MiB usage vs 942Mib with cuda 10.1 - 269MiB usage vs 2804Mib with cuda 10.2 It seems...

Hum I am not sure it will change anything as tensorrt uses cuda. Tensorrt7 requires cuda 10.2 and I may have seen an increase in memory usage (to verify) but...