Question 1

Can AMD Instinct MI250 128GB run embeddinggemma 300M?

Accepted Answer

Yes, AMD Instinct MI250 128GB can run embeddinggemma 300M with a C grade (Runs well). Expected decode speed: 2419.2 tok/s.

Question 2

How much VRAM does embeddinggemma 300M need?

Accepted Answer

embeddinggemma 300M (0.30000001192092896B parameters) requires approximately 15.0 GB of memory with Q6_K quantization.

Question 3

What is the best quantization for embeddinggemma 300M?

Accepted Answer

The recommended quantization for embeddinggemma 300M is Q6_K, which balances quality and memory efficiency.

Workload	Grade	Fit	Decode	TTFT	Context
Agentic Coding	C	Runs well	2419.2 tok/s	350 ms	272K
Chat	C	Runs well	2419.2 tok/s	350 ms	68K
Coding	C	Runs well	2419.2 tok/s	350 ms	136K
RAG	C	Runs well	2419.2 tok/s	350 ms	272K
Reasoning	C	Runs well	2419.2 tok/s	350 ms	136K

Quant	Bits	VRAM	Quality	Fit
Q2_K	2	0.1 GB	Low	D29
Q3_K_S	3	0.1 GB	Low	D29
NVFP4	4

Can AMD Instinct MI250 128GB run embeddinggemma 300M?

Memory breakdown