# Need advice on choosing model to use as movies and tv shows translator after fine-tuning on my dataset

> Source: <https://discuss.huggingface.co/t/need-advice-on-choosing-model-to-use-as-movies-and-tv-shows-translator-after-fine-tuning-on-my-dataset/191989#post_2>
> Published: 2026-11-18 12:38:26+00:00

**Recommendation:** Budget context-cache VRAM alongside model weights before choosing a model: at 128K, Mistral Small 3.2 adds about 20 GB versus Gemma 4 12B's 2.3 GB, and split long inputs into 500-800 line chunks with overlap if the card is small.
More in the playbook "Inference and self-hosting": <https://wpnews.pro/playbooks/inference-serving>

Can’t tell you which one translates Arabic or Japanese best, I haven’t tested that. But the “whole file at once” part is mostly a memory question, and there I have numbers.

Rough guess: a subtitle line is 10 to 15 tokens, so 5000 lines is somewhere around 60-75K tokens going in, and about the same coming out. So for your biggest files you need a model that can hold well over 100K tokens, and the smaller ones (200-1000 lines) fit in 32K easily.

The thing people miss is that the context itself costs memory, and how much depends heavily on the model. I keep a table of this (4-bit GGUF file + context cache), a few examples at 128K context:

Gemma 4 12B: 6.6 GB file, about 2.3 GB for the context

Gemma 4 26B-A4B: 15.8 GB file, about 2.7 GB for the context

Qwen3.8 27B: 15.3 GB file, about 8 GB for the context

Mistral Small 3.2: 13.3 GB file, about 20 GB for the context

And Qwen3 14B stops at about 40K tokens, so it’s out for the long files no matter what card you have.

So before you spend time on fine-tuning I’d check two things: the max context of the model, and what that context costs on your card. What GPU do you have? With that I can tell you which of these actually hold a 5000-line file.

One more thing, for the long files you could also split by scene (every 500-800 lines with some overlap) and pass a short summary of what happened so far. Not as clean as one pass, but it works on much smaller cards.
