Instructions to use bendavidsteel/SmolLM2-360M-Instruct-stance-target-extraction with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use bendavidsteel/SmolLM2-360M-Instruct-stance-target-extraction with PEFT:
from peft import PeftModel from transformers import AutoModelForCausalLM base_model = AutoModelForCausalLM.from_pretrained("HuggingFaceTB/SmolLM2-360M-Instruct") model = PeftModel.from_pretrained(base_model, "bendavidsteel/SmolLM2-360M-Instruct-stance-target-extraction") - Notebooks
- Google Colab
- Kaggle
Download metadata.json from bendavidsteel/SmolLM2-360M-Instruct-stance-target-extraction: direct link, hf CLI and curl.
- Browser
- Download file 1.05 kB
-
https://huggingface.co/bendavidsteel/SmolLM2-360M-Instruct-stance-target-extraction/resolve/main/metadata.json
- Command line
-
hf download hf://bendavidsteel/SmolLM2-360M-Instruct-stance-target-extraction/metadata.json
-
curl -L -o metadata.json https://huggingface.co/bendavidsteel/SmolLM2-360M-Instruct-stance-target-extraction/resolve/main/metadata.json
1.05 kB
| {"test_metrics": {"test/bertscore_f1": 0.9080705052748811, "test/bertscore_p": 0.9020601481542112, "test/bertscore_r": 0.9149142616497083, "test/bleu_f1": 66.6666666666667, "test/bleu_p": 31.415776757972356, "test/bleu_r": 35.403998937214006, "test/vast": {"bertscore_f1": 0.9153905501044131, "bertscore_p": 0.9108799733885069, "bertscore_r": 0.9207413683410819, "bleu_f1": 19.56860537594868, "bleu_p": 28.02866411631923, "bleu_r": 30.203311632457684}, "test/ezstance": {"bertscore_f1": 0.9043846915426504, "bertscore_p": 0.8976191599196459, "bertscore_r": 0.9119801771785876, "bleu_f1": 66.6666666666667, "bleu_p": 33.12126717202704, "bleu_r": 38.02266659995487}}, "prompt": "Given a piece of text, identify the stance targets - the noun phrases that the text expresses a stance on. The stance targets should be noun phrases that can be agreed or disagreed with.\nReturn only a quoted, comma-separated list of stance targets, like so: \"target1\", \"target2\"\n\nText: '{text}'\n\nStance targets:", "parent_prompt": "", "generation_method": "list"} |