AmarettoLabs/Amaretto-3B

Quantization vs mistralai/Ministral-3-3B-Instruct-2512. mistralai/Ministral-3-3B-Instruct-2512 is a quantized version of the other (F8_E4M3/BF16).

2 derivatives checked

• mradermacher/Amaretto-3B-i1-GGUF: Quantization (fewer layers)
• mradermacher/Amaretto-3B-GGUF: Same weights, converted (fewer layers)

On the Hub · Downloads on Model Pulse