Pretranslation in CafeTran is also background indexing of the current project (apart from exact and fuzzy matching) against selected memories. CafeTran lets you export the Pretranslation results to the file (Memory_Name.pretranslation) and import them back after reopening the project without the need to repeat the whole indexing process over again.
- Start a CafeTran project.
- Pretranslate it against one or more TMs (menu Translation > Pretranslate all segments)
After choosing from the menu Translation > Pretranslate all segments… you should see your opened TMs for selection (see the check boxes for each TM). No other settings are needed although you may want to free your RAM after the pretranslation is over. Then, the 'Pretranslate only' option should be selected for your TM too.
- When the Pretranslation is over, export it (menu Memory > Export > Export pretranslation… and select a destination folder). CafeTran will create a .pretranslation file for each TM in the selected folder.
- Exit CafeTran, run it again and reopen the CafeTran project.
- Select Memory > Import > Import pretranslation… and choose a .pretranslation file.
- All fuzzy, exact and fragment matches will appear instantly after each Next segment.
You can work/translate comfortably in CafeTran during Pretranslation (no need to wait until it is over). To speed and squeeze things up with large TMs.
1. Set them Read-only (it is a big RAM saver!).
2. Set them Pretranslate only (it releases RAM when the Pretranslation is over).
This (Pretranslate only setting) will also let you translate multiple TMs that in total are larger that your RAM. Just point to the folder with your TMs. Here, CafeTran loads one TM to RAM, pretranslates with it, unloads it from RAM, and takes another TM and so on.
3. Turn off the Subsegment matching for that large TMs (it will make Pretranslation much faster).
