Instructions to use EuroBERT/EuroBERT-210m with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use EuroBERT/EuroBERT-210m with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("fill-mask", model="EuroBERT/EuroBERT-210m", trust_remote_code=True)# Load model directly from transformers import AutoTokenizer, AutoModelForMaskedLM tokenizer = AutoTokenizer.from_pretrained("EuroBERT/EuroBERT-210m", trust_remote_code=True) model = AutoModelForMaskedLM.from_pretrained("EuroBERT/EuroBERT-210m", trust_remote_code=True) - Notebooks
- Google Colab
- Kaggle
Whats the closest modeling code?
#9
by michaelfeil - opened
For frameworks support, from which model should be best start modeling (llama, bert, qwen)
Hello, the easiest model to start from is Llama, as EuroBERT follows the Llama architecture, but without the causal mask.
What are other key differences besides causal=False?
Hey @michaelfeil ,our modelling is available here : https://huggingface.co/EuroBERT/EuroBERT-210m/blob/main/modeling_eurobert.py and the llama one here : https://github.com/huggingface/transformers/blob/main/src/transformers/models/llama/modeling_llama.py
It will be hard to tell you the modifications line by line. If you have a use case, feel free to share it—maybe we can help you better.
Nicolas-BZRD changed discussion status to closed