Question 1

How much VRAM does bloomz-mt need?

Accepted Answer

bloomz-mt (176.2B) requires approximately 393.94335021972654 GB of VRAM at full precision (FP16), 197 GB at 8-bit quantization (Q8), and 98.5 GB at 4-bit quantization (Q4). The actual VRAM usage may be slightly higher due to KV cache and runtime overhead.

Question 2

What GPUs can run bloomz-mt?

Accepted Answer

We currently have no compatible GPUs indexed for bloomz-mt. This may be because the model's VRAM requirements exceed available GPUs in our database, or VRAM data hasn't been computed yet.

Question 3

What is the cheapest GPU to run bloomz-mt?

Accepted Answer

We don't currently have cloud pricing data for GPUs compatible with bloomz-mt. Cloud pricing data is updated regularly — check back soon.

Question 4

Can bloomz-mt run on consumer hardware?

Accepted Answer

bloomz-mt requires approximately 393.94335021972654 GB of VRAM at FP16, and even with 4-bit quantization (Q4) it needs about 98.5 GB. This exceeds the 24 GB available on the highest-end consumer GPU (RTX 4090), so bloomz-mt requires datacenter or cloud GPUs.

Question 5

What quantization options are available for bloomz-mt?

Accepted Answer

bloomz-mt supports the following quantization levels: FP16 (full precision): ~393.94335021972654 GB, Q8 (8-bit): ~197 GB, Q6 (6-bit): ~147.7 GB, Q4 (4-bit): ~98.5 GB. Lower quantization reduces VRAM usage at the cost of some quality loss. Q8 offers a good balance between quality and memory savings, while Q4 provides maximum compression.

bloomz-mt

VRAM Requirements

Compatible GPUs

bloomz-mt FAQ

VRAM requirements for bloomz-mt at different quantization levels
Quantization	VRAM Required
FP16	393.9 GB
Q8	197 GB
Q6	147.7 GB
Q4	98.5 GB