This model is a fine-tuned version of BSC-LT/salamandra-2b. It has been trained using TRL on a Swedish version of smol-smoltalk that was translated with Gemma3-27B.

Quick start

python

from transformers import pipeline

question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
generator = pipeline("text-generation", model="liu-nlp/salamandra-2b-smol-smoltalk-sv", device="cuda")
output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
print(output["generated_text"])

Framework versions

TRL: 0.21.0
Transformers: 4.55.0
Pytorch: 2.5.1
Datasets: 4.0.0
Tokenizers: 0.21.4