french-llm-from-scratch needs about 1.3 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 8 quants, from 0.2 GB (Q4_K_M) to 1.2 GB (Q4_K_M).
French LLM from scratch, 260M parameters, Mistral‑like architecture
French‑LLM‑from‑scratch is a 260M‑parameter model trained from scratch on French conversations and FineWeb. It uses a Mistral‑like architecture with 18 layers, 16 heads, 1024‑dim embeddings and a 4096‑dim feed‑forward network. The model is released under the MIT license.
Good at
- French conversational AI
- Small‑to‑medium context handling
- Low‑resource deployment
Limits
- Context length limited to 1024 tokens
- Experimental quality, possible hallucinations
- No vision or multimodal support
Best for
- Chatting in French
- Generating French text snippets
- Fine‑tuning on French data
Who it’s for. Developers and researchers needing a lightweight French language model for experimentation or low‑resource deployment. The model is suitable for integration into local inference engines such as LM Studio or llama.cpp.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q4_K_M | 0.2 GB | 0.8 GB | Fits | Get file → | |
| Q5_K_M | 0.2 GB | 0.8 GB | Fits | Get file → | |
| Q4_K_M | 0.2 GB | 0.8 GB | Fits | Get file → | |
| Q8_0 | 0.3 GB | 0.9 GB | Fits | Get file → | |
| — | 0.7 GB | 1.2 GB | Fits | Get file → | |
| F16 | 0.7 GB | 1.3 GB | Fits | Get file → | |
| F32 | 1.2 GB | 1.7 GB | Fits | Get file → | |
| Q4_K_M | 1.2 GB | 1.7 GB | Fits | Get file → |
For 8 GB VRAM, the Q4_K_M quantized file (0.2 GB) requires about 1.3 GB of memory at 8 k context, which is well within limits. If more precision is needed and memory allows, the Q8_0 file (0.32 GB) uses about 1.4 GB. The larger F32 file (1.17 GB) needs roughly 2.3 GB and is suitable only on systems with ample VRAM.
What GPU do you need for french-llm-from-scratch?
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 Q4_K_M (1.2 GB file, about 2.3 GB in use).
Which french-llm-from-scratch quant should you choose?
Q4_K_M is the usual starting point for french-llm-from-scratch: it balances file size and quality well. Here it is a 0.2 GB file.
The Q8_0 file is 0.3 GB, about 1.6 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.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Frequently asked questions about french-llm-from-scratch
How much VRAM does french-llm-from-scratch need?
About 1.3 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 french-llm-from-scratch on an 8 GB GPU?
Yes, with Q4_K_M: the 1.2 GB file needs about 2.3 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of french-llm-from-scratch GGUF?
It depends on the quant: Q4_K_M is 0.2 GB, Q5_K_M is 0.2 GB, Q4_K_M is 0.2 GB, Q8_0 is 0.3 GB, file is 0.7 GB, F16 is 0.7 GB and 2 more.
What context length does french-llm-from-scratch support?
The GGUF header lists a trained context length of 1,024 tokens (1K). The KV cache grows with the context you use: about 0.6 GB at 8K and 0.1 GB at 1K.
Where can I download french-llm-from-scratch GGUF?
The files are on Hugging Face in the vincent-pro-ai/french-llm-from-scratch-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 french-llm-from-scratch use?
Hugging Face lists the license as MIT. 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 this model on 8 GB VRAM?
Yes, the Q4_K_M quantized version uses about 1.3 GB of VRAM at 8 k context, fitting comfortably on 8 GB.
Does the model support vision or multimodal inputs?
No, the model is purely text‑based and does not support vision or multimodal inputs.
What context length does the model support?
The model supports a maximum context length of 1024 tokens.