Mark on Batch Size

Author: lapplislazuli

README.md

@@ -95,3 +95,11 @@ CodeBert_CodeToText_Experiment_0_1  | ./entrypoint.sh: line 200: syntax error: u
 This is due to windows changing the line-breaks / file encodings. Thanks windows. 
 **Easy Solution**: run `dos2unix entrypoint.sh` and rebuild the container. 
 Its might easier/faster to pull the image from this repository, or you have to [edit the entrypoint to be compatible with windows](https://askubuntu.com/questions/966488/how-do-i-fix-r-command-not-found-errors-running-bash-scripts-in-wsl). 
+
+```
+xxx | RuntimeError: CUDA out of memory. Tried to allocate 62.00 MiB (GPU 0; 12.00 GiB total capacity; 10.57 GiB already allocated; 0 bytes free; 10.71 GiB reserved in total by PyTorch)
+```
+
+This happens in old Pytorch versions. 
+
+Reduce batch size. To the best of my knowledge, nothing else can be done about this in old pytorch versions.

docker-compose-python.yml

@@ -7,12 +7,12 @@ services:
       - ./dataset/java/:/dataset:ro
       - ./compose_output:/experiment/output
     environment:
-      epochs: 5
+      epochs: 10
       lang: python
       train_file: /dataset/train.jsonl
       valid_file: /dataset/valid.jsonl
       test_file: /dataset/test.jsonl
-      batch_size: 16
+      batch_size: 8
     oom_kill_disable: true
     deploy:
       resources: