Instructions to use toolathlonEval/Saffron-Quantization-Registry-004365 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use toolathlonEval/Saffron-Quantization-Registry-004365 with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("toolathlonEval/Saffron-Quantization-Registry-004365", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Saffron Deployment Variant Registry
This repository contains the canonical deployment packages for release SAFFRON-REGISTRY-004365.
| variant_code | candidate_key | validation_tier | build_number | package_path |
|---|---|---|---|---|
| amber-int4 | amber-int4-cert | certified | 12 | variants/amber-int4 |
| citrine-int8 | citrine-int8-r7 | verified | 7 | variants/citrine-int8 |
| topaz-fp16 | topaz-fp16-r3 | provisional | 3 | variants/topaz-fp16 |
Each package preserves the validated configuration and weight bytes from the release intake.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support