LL

llama 65b instruct

LlamaGGUF by mradermacherUpdated 2024-05-06
Download on Hugging Face
Parameters65B
Context2K
Layers80
Architecturellama
Smallest quant24.1 GB
Quick answer

llama 65b instruct needs about 61.3 GB of memory at Q4_K_M (a 39.3 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 12 quants, from 24.1 GB (Q2_K) to 46.2 GB (Q5_K_M).

LLaMA 65B Instruct – large‑scale instruction‑tuned model

LLaMA 65B Instruct is a 65‑billion‑parameter instruction‑tuned model based on the LLaMA architecture. It supports up to 2048 context tokens and is available in multiple GGUF quantizations. The model is released under a non‑commercial bespoke license.

Good at

  • Instruction following
  • Long‑form text generation
  • Large‑context handling

Limits

  • Requires high VRAM for larger quants
  • No vision support
  • License restricts commercial use

Best for

  • Enterprise internal chatbots
  • Research prototyping
  • Custom fine‑tuning

Who it’s for. Developers and researchers needing a large instruction‑tuned model for internal or research purposes. The non‑commercial license requires permission from Upstage for use.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K24.1 GB 30.0 GB
Too big Get file →
IQ3_XS26.6 GB 32.5 GB
Too big Get file →
IQ3_S28.2 GB 34.0 GB
Too big Get file →
Q3_K_S28.2 GB 34.0 GB
Too big Get file →
IQ3_M29.8 GB 35.7 GB
Too big Get file →
Q3_K_M31.6 GB 37.5 GB
Too big Get file →
Q3_K_L34.6 GB 40.5 GB
Too big Get file →
IQ4_XS35.0 GB 40.9 GB
Too big Get file →
Q4_K_S37.1 GB 42.9 GB
Too big Get file →
Q4_K_M39.3 GB 45.2 GB
Too big Get file →
Q5_K_S44.9 GB 50.8 GB
Too big Get file →
Q5_K_M46.2 GB 52.1 GB
Too big Get file →

For 8 GB VRAM, use Q2_K (file size 24.11 GB, estimated 46.1 GB memory at 8k context). For 12 GB VRAM, Q3_K_S (28.16 GB, 50.1 GB memory) is suitable. For 24 GB VRAM, Q4_K_M (39.35 GB, 61.3 GB memory) provides a balance of quality and size.

What GPU do you need for llama 65b 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: 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: no quant fits at an 8K context.
  • 32 GB VRAM: no quant fits at an 8K context.
  • 48 GB VRAM: the best quant that fits is Q2_K (24.1 GB file, about 46.1 GB in use).

Which llama 65b instruct quant should you choose?

Q4_K_M is the usual starting point for llama 65b instruct: it balances file size and quality well. Here it is a 39.3 GB file.

If memory is tight, the smallest file is Q2_K at 24.1 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
Parameters65B
Layers80
Trained context2,048 tokens
KV cache (fp16)2560 KB per token

Source

Quantized bymradermacher
FamilyLlama
Downloads (HF)1,887
Last updated2024-05-06

Frequently asked questions about llama 65b instruct

How much VRAM does llama 65b instruct need?

About 61.3 GB at Q4_K_M with an 8K context window. That is the 39.3 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 llama 65b instruct on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 46.1 GB, so you would need a larger GPU or CPU offloading.

What is the file size of llama 65b instruct GGUF?

It depends on the quant: Q2_K is 24.1 GB, IQ3_XS is 26.6 GB, IQ3_S is 28.2 GB, Q3_K_S is 28.2 GB, IQ3_M is 29.8 GB, Q3_K_M is 31.6 GB and 6 more.

What context length does llama 65b instruct support?

The GGUF header lists a trained context length of 2,048 tokens (2K). The KV cache grows with the context you use: about 21.5 GB at 8K and 5.4 GB at 2K.

Where can I download llama 65b instruct GGUF?

The files are on Hugging Face in the mradermacher/llama-65b-instruct-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Can I run LLaMA 65B Instruct on 8 GB VRAM?

Use the Q2_K quant, which requires approximately 46.1 GB of VRAM at 8k context. 8 GB VRAM is insufficient.

Does this model support image inputs?

No, the model only handles text. Image support is not available.

What license applies to this model?

It is under a non‑commercial bespoke license governed by the Meta license; commercial use is not permitted without permission.

Other GGUFs of llama-65b-instruct

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Llama models

All Llama →

Other large models

All large models →