# Can I train a model to a different downstream task?

**URL:** </static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcan-i-train-a-model-to-a-different-downstream-task%2F24385%3E%5C
**Category:** Beginners\
**Created:** [October 13, 2022, 1:50pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcan-i-train-a-model-to-a-different-downstream-task%2F24385 "2022-10-13T13:50:33Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![ypanagis](https://avatars.discourse-cdn.com/v4/letter/y/f08c70/32.png) [@ypanagis](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fypanagis)%5C
**Post date:** [October 13, 2022, 1:50pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcan-i-train-a-model-to-a-different-downstream-task%2F24385%2F1 "2022-10-13T13:50:33Z")

</div>

I am a very beginner in Transformers and HuggingFace, I searched the forum and did not find something relevant. If I understood correctly, _Transfer Learning_ should allow us to use a specific model, to new downstream tasks.

I will use a more specific example, say for example I load `bert-base-uncased`. I see that the model can be trained on eg. _Text Classification_, _Question answering_, etc. If I wanted to run an _unlisted task_, say for example NER, can I somehow train the same model, or shall I tun myself to a different model that already supports this task?

If the texts are very field specific, like science or medicine, can then `bert` be used, or shall I turn myself to more specific texts/models?

Those might be a bit detrimental to ask, but please share any information that could guide me.

PS: I have the impression that [this post](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Frelation-extraction-hugginface%2F24322) is asking a similar question.

---

<div class="post-metadata">

**Author:** ![ypanagis](https://avatars.discourse-cdn.com/v4/letter/y/f08c70/32.png) [@ypanagis](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Fu%2Fypanagis)%5C
**Post date:** [October 13, 2022, 4:20pm UTC](/static-proxy?url=https%3A%2F%2Fdiscuss.huggingface.co%2Ft%2Fcan-i-train-a-model-to-a-different-downstream-task%2F24385%2F2 "2022-10-13T16:20:01Z")

</div>

If I get it right, it is also a matter of the so called _fine-tuning_, at least if a model is trained for _Token Classification_. NER is also a form of _token classification_, thus the model has to be _adapted_.

I found relevant resources [here](https://huggingface.co/docs/transformers/v4.17.0/en/tasks/token_classification#finetune-with-trainer) for token classification and [here](https://huggingface.co/course/chapter3/3?fw=pt) for fine-tuning.