Text Classification
PEFT
Safetensors
English
biojev
biomedical
qwen3.5
qlora
natural-language-inference
biomedical-nlp
system-one
Instructions to use Gabriel382/BioJev-Nano with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use Gabriel382/BioJev-Nano with PEFT:
from peft import PeftModel from transformers import AutoModelForSequenceClassification base_model = AutoModelForSequenceClassification.from_pretrained("Qwen/Qwen3.5-0.8B-Base") model = PeftModel.from_pretrained(base_model, "Gabriel382/BioJev-Nano") - Notebooks
- Google Colab
- Kaggle
|
Download SYSTEMONE_API.md from Gabriel382/BioJev-Nano: direct link, hf CLI and curl.
- Browser
- Download file 711 Bytes
-
https://huggingface.co/Gabriel382/BioJev-Nano/resolve/main/SYSTEMONE_API.md
- Command line
-
hf download hf://Gabriel382/BioJev-Nano/SYSTEMONE_API.md
-
curl -L -o SYSTEMONE_API.md https://huggingface.co/Gabriel382/BioJev-Nano/resolve/main/SYSTEMONE_API.md
711 Bytes
| # BioJev-Nano — System One API | |
| The System One compatibility server is maintained in the main BioJev repository: | |
| https://github.com/Gabriel382/BioJev | |
| After cloning/installing BioJev, serve this local model directory with: | |
| ```bash | |
| python scripts/serve_systemone.py \ | |
| --checkpoint /path/to/BioJev-Nano \ | |
| --model-name biojev-nano \ | |
| --load-in-4bit \ | |
| --host 0.0.0.0 \ | |
| --port 8000 | |
| ``` | |
| Endpoint: | |
| ```text | |
| POST /v1/systemone | |
| ``` | |
| Supported decision types: | |
| - `choice` | |
| - `noul` | |
| - `score` | |
| BioJev-Nano remains a three-class NLI sequence classifier. The System One endpoint is | |
| a compatibility bridge which converts each candidate into an NLI hypothesis and scores | |
| it using the existing trained checkpoint. | |