MO

sapienzanlp_modello-italia-9b

ModelloGGUF by fbaldassarriUpdated 2024-07-03
Download on Hugging Face
Parameters8.54B
Context4K
Layers34
Architecturegptneox
Smallest quant3.3 GB
Quick answer

sapienzanlp_modello-italia-9b needs about 11.7 GB of memory at Q4_K_M (a 5.5 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 13 quants, from 3.3 GB (Q2_K) to 34.2 GB (F32).

Modello Italia 9B – 8.54B‑parameter GPT-NeoX model in GGUF format

Modello Italia 9B is an 8.54B‑parameter GPT‑NeoX model converted to GGUF. It supports up to 4096 context tokens and is released under the MIT license. The model is available in multiple quantizations ranging from Q2_K to F32.

Good at

  • Large context handling
  • Italian language support
  • MIT‑licensed usage

Limits

  • No vision capability
  • Requires conversion for transformers
  • No official bias mitigation

Best for

  • Chatbot development
  • Content generation in Italian
  • Research prototyping

Who it’s for. Developers and researchers who need a large Italian‑language LLM with flexible quantization options. The model is suitable for local deployment using llama.cpp or compatible frameworks.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K3.3 GB 6.6 GB
Fits Get file →
Q3_K_M4.6 GB 7.9 GB
Tight Get file →
Q4_04.9 GB 8.2 GB
Too big Get file →
Q4_K_S4.9 GB 8.3 GB
Too big Get file →
Q3_K_L4.9 GB 8.3 GB
Too big Get file →
Q4_K_M5.5 GB 8.9 GB
Too big Get file →
Q5_05.9 GB 9.3 GB
Too big Get file →
Q5_K_S5.9 GB 9.3 GB
Too big Get file →
Q5_K_M6.4 GB 9.8 GB
Too big Get file →
Q6_K7.0 GB 10.4 GB
Too big Get file →
Q8_09.1 GB 12.4 GB
Too big Get file →
F1617.1 GB 20.4 GB
Too big Get file →
F3234.2 GB 37.5 GB
Too big Get file →

For 8 GB VRAM, use Q4_K_M (5.5 GB file, ~11.7 GB memory). For 16 GB VRAM, Q5_K_M (6.41 GB file, ~12.6 GB memory) is a good balance. If more memory is available, F16 (17.08 GB file, ~23.3 GB memory) offers higher precision.

What GPU do you need for sapienzanlp_modello-italia-9b?

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: the best quant that fits is Q4_K_M (5.5 GB file, about 11.7 GB in use).
  • 16 GB VRAM: the best quant that fits is Q8_0 (9.1 GB file, about 15.3 GB in use).
  • 24 GB VRAM: the best quant that fits is F16 (17.1 GB file, about 23.3 GB in use).
  • 32 GB VRAM: the best quant that fits is F16 (17.1 GB file, about 23.3 GB in use).
  • 48 GB VRAM: the best quant that fits is F32 (34.2 GB file, about 40.4 GB in use).

Which sapienzanlp_modello-italia-9b quant should you choose?

Q4_K_M is the usual starting point for sapienzanlp_modello-italia-9b: it balances file size and quality well. Here it is a 5.5 GB file.

The Q8_0 file is 9.1 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 3.3 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

Architecturegptneox
Parameters8.54B
Layers34
Trained context4,096 tokens
KV cache (fp16)680 KB per token

Source

Quantized byfbaldassarri
FamilyModello
Licensemit
Downloads (HF)627
Last updated2024-07-03

Frequently asked questions about sapienzanlp_modello-italia-9b

How much VRAM does sapienzanlp_modello-italia-9b need?

About 11.7 GB at Q4_K_M with an 8K context window. That is the 5.5 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 sapienzanlp_modello-italia-9b on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 9.5 GB, so you would need a larger GPU or CPU offloading.

What is the file size of sapienzanlp_modello-italia-9b GGUF?

It depends on the quant: Q2_K is 3.3 GB, Q3_K_M is 4.6 GB, Q4_0 is 4.9 GB, Q4_K_S is 4.9 GB, Q3_K_L is 4.9 GB, Q4_K_M is 5.5 GB and 7 more.

What context length does sapienzanlp_modello-italia-9b support?

The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 5.7 GB at 8K and 2.9 GB at 4K.

Where can I download sapienzanlp_modello-italia-9b GGUF?

The files are on Hugging Face in the fbaldassarri/modello-italia-9B-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 sapienzanlp_modello-italia-9b 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 Modello Italia 9B on 8 GB VRAM?

Yes, the Q4_K_M quantization requires about 11.7 GB of VRAM at 8 k context, which fits within an 8 GB GPU with some overhead.

Does the model support Italian only?

The model is trained primarily on Italian data, but it can handle other languages to a limited extent.

Is the model free to use?

Yes, it is released under the MIT license, allowing free use and modification.

Other GGUFs of modello-italia-9b

Similar models

Browse all →

More Modello models

All Modello →

Other small models

All small models →