4-bit Open on Hugging Face
Parameters
1.4B
Weights (est.)
≈ 790 MB
Mac memory (est.)
8 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Details
- Base model
- openai/privacy-filter (finetune)
- Vendor
- OpenAI · gpt-oss
- Task
- Token classification
- Quantisation
- 4-bit
- Licence
- apache-2.0
- Downloads
- 413 in the last 30 days · 4.4K all time · 6 likes
- Published
- 2026-04-22 · updated 2026-04-23 (6mo ago)
- Tags
- openai_privacy_filter, transformers.js, mlx-embeddings
Other MLX builds of openai/privacy-filter
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| openai-privacy-filter-4bit (this one) | 4-bit | ≈ 790 MB | 8 GB+ | 413 |
| openai-privacy-filter-mxfp4 | MXFP4 | ≈ 790 MB | 8 GB+ | 186 |
| openai-privacy-filter-5bit | 5-bit | ≈ 965 MB | 8 GB+ | 162 |
| openai-privacy-filter-6bit | 6-bit | ≈ 1.1 GB | 8 GB+ | 168 |
| openai-privacy-filter-8bit | 8-bit | ≈ 1.5 GB | 8 GB+ | 267 |
| openai-privacy-filter-mxfp8 | MXFP8 | ≈ 1.5 GB | 8 GB+ | 219 |
| openai-privacy-filter-bf16 | bf16 | ≈ 2.8 GB | 8 GB+ | 420 |
| openai-privacy-filter-nvfp4 | — | ≈ 5.6 GB | 16 GB+ | 190 |