Sarvam 30B ROCmFP4 needs about 18.3 GB of memory at Q4_0 (a 17.6 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 4 quants, from 17.4 GB (Q4_0) to 33.7 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 0.6 GB.
Sarvam 30B ROCmFP4 is a 32.2B-parameter Sarvam model in GGUF format, quantized by kingjones777. Its trained context length is 128K tokens. The repository has 4 quant files, from 17.4 GB (Q4_0) to 33.7 GB (Q8_0).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_0 | 17.4 GB | 18.1 GB | Too big | Get file → | |
| Q4_0 | 17.6 GB | 18.3 GB | Too big | Get file → | |
| Q4_0 | 18.7 GB | 19.3 GB | Too big | Get file → | |
| Q8_0 | 33.7 GB | 34.4 GB | Too big | Get file → |
What GPU do you need for Sarvam 30B ROCmFP4?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: no quant fits at an 8K context.
- 6 GB VRAM: no quant fits at an 8K context.
- 8 GB VRAM: no quant fits at an 8K context.
- 12 GB VRAM: no quant fits at an 8K context.
- 16 GB VRAM: no quant fits at an 8K context.
- 24 GB VRAM: the best quant that fits is Q4_0 (18.7 GB file, about 19.3 GB in use).
- 32 GB VRAM: the best quant that fits is Q4_0 (18.7 GB file, about 19.3 GB in use).
- 48 GB VRAM: the best quant that fits is Q8_0 (33.7 GB file, about 34.4 GB in use).
Which Sarvam 30B ROCmFP4 quant should you choose?
Q4_0 (17.6 GB) is the middle option in this repository. The usual sweet spot between file size and quality, and the quant most people start with.
If memory is tight, the smallest file is Q4_0 at 17.4 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Sarvam 30B ROCmFP4
How much VRAM does Sarvam 30B ROCmFP4 need?
About 18.3 GB at Q4_0 with an 8K context window. That is the 17.6 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Sarvam 30B ROCmFP4 on an 8 GB GPU?
Not at an 8K context. The smallest file needs about 18.1 GB, so you would need a larger GPU or CPU offloading.
What is the file size of Sarvam 30B ROCmFP4 GGUF?
It depends on the quant: Q4_0 is 17.4 GB, Q4_0 is 17.6 GB, Q4_0 is 18.7 GB, Q8_0 is 33.7 GB.
What context length does Sarvam 30B ROCmFP4 support?
The GGUF header lists a trained context length of 131,072 tokens (128K). The KV cache grows with the context you use: about 0.2 GB at 8K and 0.6 GB at 32K.
Where can I download Sarvam 30B ROCmFP4 GGUF?
The files are on Hugging Face in the kingjones777/Sarvam-30B-ROCmFP4-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Sarvam 30B ROCmFP4 use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.