Salamandra-7B-Instruct-MLX-4bit

MLX 4-bit quantisation of BSC-LT/salamandra-7b-instruct, converted for use on Apple Silicon via mlx-lm.

Source model

• Repository: BSC-LT/salamandra-7b-instruct
• Release: 2024-09
• Family: salamandra
• Origin: eu
• Languages / coverage: 35 European languages, multilingual from-scratch pretraining at BSC MareNostrum 5
• License: apache-2.0 (inherited)

Notes from upstream

Barcelona Supercomputing Center. Part of the Salamandra family (2B / 7B / 40B). Apache 2.0.

Conversion details

• Tool: mlx-lm 0.31.3
• Quantisation: 4-bit (defaults from mlx_lm.convert)
• Converted on: 2026-05-05

Usage

from mlx_lm import load, generate

model, tokenizer = load("luiscalisto/Salamandra-7B-Instruct-MLX-4bit")
prompt = "Hello, who are you?"
print(generate(model, tokenizer, prompt=prompt, max_tokens=128, verbose=False))

License and attribution

This is a quantised redistribution of BSC-LT/salamandra-7b-instruct. The original model and its license terms (apache-2.0) carry through unchanged. Please cite the upstream authors when using this model. See the source repository for the authoritative model card and citation.

Conversion provenance

Produced by llm-mlx-conversions, a small utility for publishing community MLX 4-bit quants of open-weight LLMs.