Hi Fair Quokka, You are right. The "libtglang.so" isn't pointing to the right path of my ML model (libnimlib.so). I'm attaching the fixed "libtglang.so".

Here is the full tree as an example (from libtglang-tester).
.
├── CMakeLists.txt
├── build
│ └── tglang-tester
├── libtglang.so <= FIXED
├── resources
│ └── libnimlib.so <= ML MODEL
├── tglang-tester.c
└── tglang.h

Here is whole submission folder, but feel free to only replace the "libtglang.so" file.

https://drive.google.com/drive/folders/1QosZXGL1Vj0wnHDEU2bVLXTQiukYxggJ?usp=sharing

I know that I might not win due to this mistake. But I would request to you to test my results. I followed a very different approach than neural networks or related models.

Thank you so much! It was a fun contest!

In our final tests, this submission received the following scores (out of 100):

Code/Other Detection: 42.8
Language Detection: 0
Average Time: 19ms per sample

Final Score: 0

This is an important message regarding your submission. The following issues have been discovered during preliminary testing:
error while loading shared libraries: libnimlib.so: cannot open shared object file: No such file or directory

You can provide further clarifications within the next 24 hours.