Instructions to use DDSC/roberta-base-danish with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use DDSC/roberta-base-danish with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("fill-mask", model="DDSC/roberta-base-danish")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForMaskedLM tokenizer = AutoTokenizer.from_pretrained("DDSC/roberta-base-danish") model = AutoModelForMaskedLM.from_pretrained("DDSC/roberta-base-danish", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Download makefile from DDSC/roberta-base-danish: direct link, hf CLI and curl.
- Browser
- Download file 462 Bytes
-
https://huggingface.co/DDSC/roberta-base-danish/resolve/main/makefile
- Command line
-
hf download hf://DDSC/roberta-base-danish/makefile
-
curl -L -o makefile https://huggingface.co/DDSC/roberta-base-danish/resolve/main/makefile
462 Bytes
| train: | |
| python3 ./src/run_mlm_flax.py \ | |
| --output_dir="." \ | |
| --model_type="roberta" \ | |
| --config_name="." \ | |
| --tokenizer_name="." \ | |
| --max_seq_length="128" \ | |
| --weight_decay="0.01" \ | |
| --per_device_train_batch_size="128" \ | |
| --per_device_eval_batch_size="128" \ | |
| --learning_rate="3e-4" \ | |
| --warmup_steps="1000" \ | |
| --overwrite_output_dir \ | |
| --pad_to_max_length \ | |
| --num_train_epochs="18" \ | |
| --adam_beta1="0.9" \ | |
| --adam_beta2="0.98" \ | |
| --push_to_hub | |