LL

CodeLlama 70b Instruct iMat

LlamaGGUF by dranger003Updated 2024-02-19
Download on Hugging Face
Parameters69B
Context4K
Layers80
Architecturellama
Smallest quant14.5 GB
Quick answer

CodeLlama 70b Instruct iMat needs about 21.5 GB of memory at IQ2_XXS (a 18.3 GB file) with an 8K context window, so it fits on a 24 GB GPU. This repository has 4 quants, from 14.5 GB (IQ1_S) to 27.0 GB (IQ3_XXS).

CodeLlama 70b Instruct iMat is a 69B-parameter Llama model in GGUF format, quantized by dranger003. Its trained context length is 4K tokens. The repository has 4 quant files, from 14.5 GB (IQ1_S) to 27.0 GB (IQ3_XXS).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ1_S14.5 GB 16.4 GB
Too big Get file →
IQ2_XXS18.3 GB 20.1 GB
Too big Get file →
IQ2_XS20.3 GB 22.2 GB
Too big Get file →
IQ3_XXS27.0 GB 28.9 GB
Too big Get file →

What GPU do you need for CodeLlama 70b Instruct iMat?

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: no quant fits at an 8K context.
  • 16 GB VRAM: no quant fits at an 8K context.
  • 24 GB VRAM: the best quant that fits is IQ2_XS (20.3 GB file, about 23.5 GB in use).
  • 32 GB VRAM: the best quant that fits is IQ3_XXS (27.0 GB file, about 30.2 GB in use).

Which CodeLlama 70b Instruct iMat quant should you choose?

IQ2_XXS (18.3 GB) is the middle option in this repository. 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.

If memory is tight, the smallest file is IQ1_S at 14.5 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

Architecturellama
Parameters69B
Layers80
Trained context4,096 tokens
KV cache (fp16)320 KB per token

Source

Quantized bydranger003
FamilyLlama
Licensellama2
Downloads (HF)77
Last updated2024-02-19

Frequently asked questions about CodeLlama 70b Instruct iMat

How much VRAM does CodeLlama 70b Instruct iMat need?

About 21.5 GB at IQ2_XXS with an 8K context window. That is the 18.3 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 CodeLlama 70b Instruct iMat on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 17.7 GB, so you would need a larger GPU or CPU offloading.

What is the file size of CodeLlama 70b Instruct iMat GGUF?

It depends on the quant: IQ1_S is 14.5 GB, IQ2_XXS is 18.3 GB, IQ2_XS is 20.3 GB, IQ3_XXS is 27.0 GB.

What context length does CodeLlama 70b Instruct iMat support?

The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 2.7 GB at 8K and 1.3 GB at 4K.

Where can I download CodeLlama 70b Instruct iMat GGUF?

The files are on Hugging Face in the dranger003/CodeLlama-70b-Instruct-iMat.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 CodeLlama 70b Instruct iMat use?

Hugging Face lists the license as llama2. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Similar models

Browse all →

More from dranger003

All by dranger003 →

More Llama models

All Llama →

Other large models

All large models →