LL

CodeLlama 7B Instruct

LlamaGGUF by SinpxAIUpdated 2024-03-06
Download on Hugging Face
Parameters6.74B
Context16K
Layers32
Architecturellama
Smallest quant2.5 GB
Quick answer

CodeLlama 7B Instruct needs about 8.9 GB of memory at Q4_K_M (a 4.1 GB file) with an 8K context window, so it fits on a 12 GB GPU. This repository has 12 quants, from 2.5 GB (Q2_K) to 7.2 GB (Q8_0).

CodeLlama 7B Instruct is a 6.74B-parameter Llama model in GGUF format, quantized by SinpxAI. Its trained context length is 16K tokens. The repository has 12 quant files, from 2.5 GB (Q2_K) to 7.2 GB (Q8_0).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K2.5 GB 7.3 GB
Tight Get file →
Q3_K_S2.9 GB 7.7 GB
Tight Get file →
Q3_K_M3.3 GB 8.1 GB
Too big Get file →
Q3_K_L3.6 GB 8.4 GB
Too big Get file →
Q4_03.8 GB 8.6 GB
Too big Get file →
Q4_K_S3.9 GB 8.7 GB
Too big Get file →
Q4_K_M4.1 GB 8.9 GB
Too big Get file →
Q5_04.7 GB 9.4 GB
Too big Get file →
Q5_K_S4.7 GB 9.4 GB
Too big Get file →
Q5_K_M4.8 GB 9.6 GB
Too big Get file →
Q6_K5.5 GB 10.3 GB
Too big Get file →
Q8_07.2 GB 12.0 GB
Too big Get file →

What GPU do you need for CodeLlama 7B Instruct?

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: the best quant that fits is Q3_K_S (2.9 GB file, about 7.7 GB in use).
  • 12 GB VRAM: the best quant that fits is Q8_0 (7.2 GB file, about 12.0 GB in use).

Which CodeLlama 7B Instruct quant should you choose?

Q4_K_M is the usual starting point for CodeLlama 7B Instruct: it balances file size and quality well. Here it is a 4.1 GB file.

The Q8_0 file is 7.2 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.

If memory is tight, the smallest file is Q2_K at 2.5 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
Parameters6.74B
Layers32
Trained context16,384 tokens
KV cache (fp16)512 KB per token

Source

Quantized bySinpxAI
FamilyLlama
Downloads (HF)135
Last updated2024-03-06

Frequently asked questions about CodeLlama 7B Instruct

How much VRAM does CodeLlama 7B Instruct need?

About 8.9 GB at Q4_K_M with an 8K context window. That is the 4.1 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 7B Instruct on an 8 GB GPU?

Yes, with Q3_K_S: the 2.9 GB file needs about 7.7 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of CodeLlama 7B Instruct GGUF?

It depends on the quant: Q2_K is 2.5 GB, Q3_K_S is 2.9 GB, Q3_K_M is 3.3 GB, Q3_K_L is 3.6 GB, Q4_0 is 3.8 GB, Q4_K_S is 3.9 GB and 6 more.

What context length does CodeLlama 7B Instruct support?

The GGUF header lists a trained context length of 16,384 tokens (16K). The KV cache grows with the context you use: about 4.3 GB at 8K and 8.6 GB at 16K.

Where can I download CodeLlama 7B Instruct GGUF?

The files are on Hugging Face in the SinpxAI/CodeLlama-7B-Instruct-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Similar models

Browse all →

More from SinpxAI

All by SinpxAI →

More Llama models

All Llama →

Other small models

All small models →