bigcode-starcoder2-15b-4bit
mlx-community/bigcode-starcoder2-15b-4bit
Text generation · transformers
4-bit Open on Hugging Face
Parameters
16B
Weights (est.)
≈ 9.0 GB
Mac memory (est.)
16 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/bigcode-starcoder2-15b-4bit --prompt "Hello"
Details
- Base model
- bigcode/starcoder2-15b (quantized)
- Vendor
- BigCode
- Task
- Text generation
- Quantisation
- 4-bit
- Licence
- bigcode-openrail-m
- Downloads
- 198 in the last 30 days · 1.4K all time · 0 likes
- Published
- 2025-02-13 · updated 2025-02-16 (1y ago)
- Tags
- starcoder2, code, model-index
Other MLX builds of bigcode/starcoder2-15b
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| bigcode-starcoder2-15b-4bit (this one) | 4-bit | ≈ 9.0 GB | 16 GB+ | 198 |
| starcoder2-15b-instruct-v0.1-4bit | 4-bit | ≈ 9.0 GB | 16 GB+ | 108 |
| bigcode-starcoder2-15b-6bit | 6-bit | ≈ 13.0 GB | 24 GB+ | 80 |
| bigcode-starcoder2-15b-8bit | 8-bit | ≈ 17.0 GB | 24 GB+ | 96 |
| starcoder2-15b-instruct-v0.1-8bit | 8-bit | ≈ 17.0 GB | 24 GB+ | 75 |
| starcoder2-15b-instruct-v0.1 | — | ≈ 31.9 GB | 48 GB+ | 96 |
| bigcode-starcoder2-15b-fp16 | fp16 | ≈ 31.9 GB | 48 GB+ | 86 |
| bigcode-starcoder2-15b-fp32 | fp32 | ≈ 63.8 GB | 96 GB+ | 186 |