SatDiff LFM2.5 VL 450M stage1 needs about 1.4 GB of memory at Q8_0 (a 0.4 GB file) with an 8K context window, so it fits on a 4 GB GPU. Longer context needs more: at 32K tokens the KV cache alone adds about 2.1 GB.
SatDiff LFM2.5 VL 450M stage1 is a 354M-parameter Lfm model in GGUF format, quantized by WobblyDopamine. Its trained context length is 125K tokens. The repository has 1 quant file, from 0.4 GB (Q8_0) to 0.4 GB (Q8_0).
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q8_0 | 0.4 GB | 1.4 GB | Fits | Get file → |
What GPU do you need for SatDiff LFM2.5 VL 450M stage1?
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: the best quant that fits is Q8_0 (0.4 GB file, about 1.4 GB in use).
Which SatDiff LFM2.5 VL 450M stage1 quant should you choose?
Q8_0 (0.4 GB) is the middle option in this repository. Close to lossless. It costs roughly twice the memory of a 4-bit file, so it suits cases where you have plenty of VRAM.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about SatDiff LFM2.5 VL 450M stage1
How much VRAM does SatDiff LFM2.5 VL 450M stage1 need?
About 1.4 GB at Q8_0 with an 8K context window. That is the 0.4 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 SatDiff LFM2.5 VL 450M stage1 on an 8 GB GPU?
Yes, with Q8_0: the 0.4 GB file needs about 1.4 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of SatDiff LFM2.5 VL 450M stage1 GGUF?
It depends on the quant: Q8_0 is 0.4 GB.
What context length does SatDiff LFM2.5 VL 450M stage1 support?
The GGUF header lists a trained context length of 128,000 tokens (125K). The KV cache grows with the context you use: about 0.5 GB at 8K and 2.1 GB at 32K.
Where can I download SatDiff LFM2.5 VL 450M stage1 GGUF?
The files are on Hugging Face in the WobblyDopamine/SatDiff-LFM2.5-VL-450M-stage1 repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does SatDiff LFM2.5 VL 450M stage1 use?
Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.