4-bit Open on Hugging Face
Parameters
33B
Weights (est.)
≈ 18.4 GB
Mac memory (est.)
24 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.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/SERA-32B-GA-4bit --prompt "Hello"
Details
- Base model
- allenai/SERA-32B-GA (quantized)
- Vendor
- Ai2 · OLMo
- Task
- Text generation
- Quantisation
- 4-bit
- Licence
- apache-2.0
- Downloads
- 20 in the last 30 days · 133 all time · 1 likes
- Published
- 2026-01-27 · updated 2026-01-27 (9mo ago)
- Tags
- qwen3
Other MLX builds of allenai/SERA-32B-GA
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| SERA-32B-GA-4bit (this one) | 4-bit | ≈ 18.4 GB | 24 GB+ | 20 |
| SERA-32B-GA-5bit | 5-bit | ≈ 22.5 GB | 32 GB+ | 11 |
| SERA-32B-GA-6bit | 6-bit | ≈ 26.6 GB | 32 GB+ | 12 |
| SERA-32B-GA-8bit | 8-bit | ≈ 34.8 GB | 48 GB+ | 12 |
| SERA-32B-GA-bf16 | bf16 | ≈ 65.5 GB | 96 GB+ | 10 |