Summarization
Transformers
Safetensors
Latvian
mbart
text2text-generation
mbart50
low-resource
research
rahvusarhiiv
Eval Results (legacy)
Instructions to use Rahvusarhiiv/lv_summariser with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Rahvusarhiiv/lv_summariser with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "summarization" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # 'pip install "transformers<5.0.0' from transformers import pipeline pipe = pipeline("summarization", model="Rahvusarhiiv/lv_summariser")# Load model directly from transformers import AutoTokenizer, AutoModelForSeq2SeqLM tokenizer = AutoTokenizer.from_pretrained("Rahvusarhiiv/lv_summariser") model = AutoModelForSeq2SeqLM.from_pretrained("Rahvusarhiiv/lv_summariser", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| coverage,dataset,eu_legalese_template_flag,examples,factuality,focus,hit_max_new_tokens,kind,label,language_score,overall,prediction_source_content_token_precision,prediction_words,reference_content_token_coverage,repeated_4gram_rate,rouge1_f,rouge2_f,rougeL_f | |
| ,wiki4_lv,0.0,100,,,0.0,automatic,Strict Wiki4 automatic,,,0.8635543226625836,28.73,0.2909341006437182,0.0,0.34677778028811274,0.2513317175742254,0.31327548240008685 | |
| ,gdelt_lv,0.0,100,,,0.0,automatic,Recent GDELT news automatic,,,0.8133434324159858,20.97,0.2170002081154365,0.0,0.23995412053532847,0.10362259282369267,0.20197108972197333 | |
| ,news_supported_lv,0.0,100,,,0.0,automatic,Held-out supported news automatic,,,0.8163536151873175,24.37,0.283575482256744,0.0,0.3308161717084071,0.17596543196643705,0.28414998231738237 | |
| 2.2,wiki4_lv,,5,3.8,4.4,,ollama_judge,Strict Wiki4 paired LLM judge,5.0,2.6,,,,,,, | |
| 2.8,gdelt_lv,,5,2.8,4.0,,ollama_judge,Recent GDELT news paired LLM judge,5.0,2.4,,,,,,, | |
| 2.4,news_supported_lv,,5,3.2,4.0,,ollama_judge,Held-out supported news paired LLM judge,4.8,2.6,,,,,,, | |