MI

kolibri mistral 0427 2

MistralGGUF by kaleinaNyanUpdated 2024-05-13
Download on Hugging Face
Parameters7.24B
Context32K
Layers32
Architecturellama
Smallest quant4.4 GB
Quick answer

kolibri mistral 0427 2 needs about 5.9 GB of memory at Q4_K_M (a 4.4 GB file) with an 8K context window, so it fits on a 6 GB GPU. This repository has 2 quants, from 4.4 GB (Q4_K_M) to 7.7 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 4.3 GB.

Instruction following model optimized for Russian language

Kolibri Mistral 0427 2 is an instruction following model based on Mistral v0.1 Instruct. It is optimized for Russian language and trained using kolibrify on a variety of instruction datasets. The model uses the ChatML template and is recommended for use with LMStudio.

Good at

  • Russian instruction following
  • ChatML template support
  • Mistral architecture

Limits

  • No vision support
  • Pre-alpha release
  • License apache-2.0

Best for

  • Russian language chatbots
  • Instruction following tasks
  • LMStudio experimentation

Who it’s for. Developers building Russian LLM applications and researchers testing instruction following can use this model. It is suitable for early experimentation and prototyping.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q4_K_M4.4 GB 5.9 GB
Fits Get file →
Q8_07.7 GB 9.3 GB
Too big Get file →

For 8 GB VRAM, use the Q4_K_M quant which is 4.37 GB and requires about 5.9 GB at 8k context. If you have more memory, the Q8_0 quant is 7.7 GB and needs about 9.3 GB at 8k context. These estimates are approximate.

What GPU do you need for kolibri mistral 0427 2?

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: the best quant that fits is Q4_K_M (4.4 GB file, about 5.9 GB in use).
  • 8 GB VRAM: the best quant that fits is Q4_K_M (4.4 GB file, about 5.9 GB in use).
  • 12 GB VRAM: the best quant that fits is Q8_0 (7.7 GB file, about 9.3 GB in use).

Which kolibri mistral 0427 2 quant should you choose?

Q4_K_M is the usual starting point for kolibri mistral 0427 2: it balances file size and quality well. Here it is a 4.4 GB file.

The Q8_0 file is 7.7 GB, about 1.8 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

Architecturellama
Parameters7.24B
Layers32
Trained context32,768 tokens
KV cache (fp16)128 KB per token

Source

Quantized bykaleinaNyan
FamilyMistral
Licenseapache-2.0
Downloads (HF)1,465
Last updated2024-05-13

Frequently asked questions about kolibri mistral 0427 2

How much VRAM does kolibri mistral 0427 2 need?

About 5.9 GB at Q4_K_M with an 8K context window. That is the 4.4 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 kolibri mistral 0427 2 on an 8 GB GPU?

Yes, with Q4_K_M: the 4.4 GB file needs about 5.9 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of kolibri mistral 0427 2 GGUF?

It depends on the quant: Q4_K_M is 4.4 GB, Q8_0 is 7.7 GB.

What context length does kolibri mistral 0427 2 support?

The GGUF header lists a trained context length of 32,768 tokens (32K). The KV cache grows with the context you use: about 1.1 GB at 8K and 4.3 GB at 32K.

Where can I download kolibri mistral 0427 2 GGUF?

The files are on Hugging Face in the kaleinaNyan/kolibri-mistral-0427-2.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 kolibri mistral 0427 2 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 this model on 8 GB VRAM?

Yes, the Q4_K_M quant uses about 5.9 GB at 8k context, fitting in 8 GB VRAM.

Does it support vision?

No, this model has no vision capability.

What license does it use?

It is released under the Apache-2.0 license.

Similar models

Browse all →

More from kaleinaNyan

All by kaleinaNyan →

More Mistral models

All Mistral →

Other small models

All small models →