FinxGemma Lite needs about 2.3 GB of memory at Q4_K_M (a 0.2 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 12 quants, from 0.1 GB (Q2_K) to 0.6 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 6.4 GB.
FinxGemma Lite – a 0.271B embedding model fine‑tuned for transaction descriptors
FinxGemma Lite is a 0.271B parameter embedding model based on gemma-embedding2. It is fine‑tuned on raw card‑transaction merchant descriptors for retrieval‑based transaction categorisation. The model is released under the apache‑2.0 license.
Good at
- Embedding short, noisy transaction strings
- Retrieval‑based merchant categorisation
- Low‑memory inference
Limits
- No vision support
- Not a general‑purpose sentence embedding
- Requires a confidence layer for financial decisions
Best for
- Embedding new transaction descriptors
- Nearest‑neighbour lookup of known merchants
- Research prototyping of retrieval systems
Who it’s for. Developers and researchers building retrieval‑based categorisation systems for card and bank transaction descriptors. The model is suitable for prototyping and research, not for automated financial decisions without a confidence layer.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 0.1 GB | 2.2 GB | Fits | Get file → | |
| Q3_K_S | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q3_K_M | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q3_K_L | 0.2 GB | 2.3 GB | Fits | Get file → | |
| IQ4_XS | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q4_K_S | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q4_K_M | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q5_K_S | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q5_K_M | 0.2 GB | 2.3 GB | Fits | Get file → | |
| Q6_K | 0.3 GB | 2.4 GB | Fits | Get file → | |
| Q8_0 | 0.3 GB | 2.4 GB | Fits | Get file → | |
| F16 | 0.6 GB | 2.7 GB | Fits | Get file → |
For 8 GB VRAM, Q4_K_M (0.19 GB) or Q5_K_M (0.23 GB) are suitable, using approximately 2.3 GB of VRAM at 8 k context. For tighter budgets, Q2_K (0.13 GB) uses about 2.2 GB of VRAM. Larger quants such as F16 (0.57 GB) require more memory but provide higher precision.
What GPU do you need for FinxGemma Lite?
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 F16 (0.6 GB file, about 2.7 GB in use).
Which FinxGemma Lite quant should you choose?
Q4_K_M is the usual starting point for FinxGemma Lite: it balances file size and quality well. Here it is a 0.2 GB file.
The Q8_0 file is 0.3 GB, about 1.7 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q2_K at 0.1 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about FinxGemma Lite
How much VRAM does FinxGemma Lite need?
About 2.3 GB at Q4_K_M with an 8K context window. That is the 0.2 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 FinxGemma Lite on an 8 GB GPU?
Yes, with F16: the 0.6 GB file needs about 2.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of FinxGemma Lite GGUF?
It depends on the quant: Q2_K is 0.1 GB, Q3_K_S is 0.2 GB, Q3_K_M is 0.2 GB, Q3_K_L is 0.2 GB, IQ4_XS is 0.2 GB, Q4_K_S is 0.2 GB and 6 more.
What context length does FinxGemma Lite support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 1.6 GB at 8K and 6.4 GB at 32K.
Where can I download FinxGemma Lite GGUF?
The files are on Hugging Face in the mradermacher/FinxGemma-Lite-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 FinxGemma Lite 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.
Can I run FinxGemma Lite on 8 GB VRAM?
Yes. Quants such as Q4_K_M or Q5_K_M fit within 8 GB VRAM, using roughly 2.3 GB at 8 k context.
Does the model support image input?
No. FinxGemma Lite is an embedding model and does not support vision.
Is the model suitable for production financial decisions?
No. It is intended for research and prototyping; a confidence layer is required for production use.