AI

bhenrym14_airoboros-l2-13b-2.1-yarn-64k

AiroborosGGUF by TheBlokeUpdated 2023-09-27
Download on Hugging Face
Parameters13B
Context64K
Layers40
Architecturellama
Smallest quant5.4 GB
Quick answer

bhenrym14_airoboros-l2-13b-2.1-yarn-64k needs about 15.1 GB of memory at Q4_K_M (a 7.9 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 12 quants, from 5.4 GB (Q2_K) to 13.8 GB (Q8_0). Longer context needs more: at 32K tokens the KV cache alone adds about 26.8 GB.

13B Llama‑2 model with 64k token context via YaRN

This is a finetune of Llama‑2‑13b with YaRN scaling to extend context to 64k tokens. It is instruction tuned on the Airoboros 2.1 dataset using QLoRA rank 64. The model is available in GGUF format with multiple quantisations.

Good at

  • Long‑context instruction following
  • YaRN‑scaled RoPE performance
  • LLM inference on local hardware

Limits

  • Requires large VRAM for full 64k context
  • No native flash attention support
  • License is llama2

Best for

  • Chat applications needing extended context
  • Research on RoPE scaling
  • Local LLM deployment with GGUF

Who it’s for. Developers and researchers who need a 13B Llama‑2 model that can handle up to 64k tokens. The model is suitable for local inference on machines with sufficient GPU memory and for testing YaRN scaling effects.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K5.4 GB 12.6 GB
Too big Get file →
Q3_K_S5.7 GB 12.9 GB
Too big Get file →
Q3_K_M6.3 GB 13.5 GB
Too big Get file →
Q3_K_L6.9 GB 14.1 GB
Too big Get file →
Q4_07.4 GB 14.6 GB
Too big Get file →
Q4_K_S7.4 GB 14.6 GB
Too big Get file →
Q4_K_M7.9 GB 15.1 GB
Too big Get file →
Q5_09.0 GB 16.2 GB
Too big Get file →
Q5_K_S9.0 GB 16.2 GB
Too big Get file →
Q5_K_M9.2 GB 16.4 GB
Too big Get file →
Q6_K10.7 GB 17.9 GB
Too big Get file →
Q8_013.8 GB 21.0 GB
Too big Get file →

For 8 GB VRAM, use Q4_K or Q4_K_S (file size 7.37 GB or 7.41 GB, estimated 14.6 GB memory at 8k context). For 12 GB VRAM, Q5_K_S or Q5_K_M (file size 8.97 GB or 9.23 GB, estimated 16.2 GB or 16.4 GB). For 16 GB VRAM, Q6_K (file size 10.68 GB, estimated 17.9 GB).

What GPU do you need for bhenrym14_airoboros-l2-13b-2.1-yarn-64k?

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: the best quant that fits is Q4_K_M (7.9 GB file, about 15.1 GB in use).
  • 24 GB VRAM: the best quant that fits is Q8_0 (13.8 GB file, about 21.0 GB in use).

Which bhenrym14_airoboros-l2-13b-2.1-yarn-64k quant should you choose?

Q4_K_M is the usual starting point for bhenrym14_airoboros-l2-13b-2.1-yarn-64k: it balances file size and quality well. Here it is a 7.9 GB file.

The Q8_0 file is 13.8 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 5.4 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
Parameters13B
Layers40
Trained context65,536 tokens
KV cache (fp16)800 KB per token

Source

Quantized byTheBloke
FamilyAiroboros
Licensellama2
Downloads (HF)2,711
Last updated2023-09-27

Frequently asked questions about bhenrym14_airoboros-l2-13b-2.1-yarn-64k

How much VRAM does bhenrym14_airoboros-l2-13b-2.1-yarn-64k need?

About 15.1 GB at Q4_K_M with an 8K context window. That is the 7.9 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 bhenrym14_airoboros-l2-13b-2.1-yarn-64k on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 12.6 GB, so you would need a larger GPU or CPU offloading.

What is the file size of bhenrym14_airoboros-l2-13b-2.1-yarn-64k GGUF?

It depends on the quant: Q2_K is 5.4 GB, Q3_K_S is 5.7 GB, Q3_K_M is 6.3 GB, Q3_K_L is 6.9 GB, Q4_0 is 7.4 GB, Q4_K_S is 7.4 GB and 6 more.

What context length does bhenrym14_airoboros-l2-13b-2.1-yarn-64k support?

The GGUF header lists a trained context length of 65,536 tokens (64K). The KV cache grows with the context you use: about 6.7 GB at 8K and 26.8 GB at 32K.

Where can I download bhenrym14_airoboros-l2-13b-2.1-yarn-64k GGUF?

The files are on Hugging Face in the TheBloke/Airoboros-L2-13B-2_1-YaRN-64K-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 bhenrym14_airoboros-l2-13b-2.1-yarn-64k 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.

Can I run this model on 8 GB VRAM?

Use Q4_K or Q4_K_S quantisation; the estimated VRAM usage is about 14.6 GB at 8k context, which is above 8 GB, so it will not fit without further optimisation.

Does this model support vision?

No, the model is purely text and does not include vision capabilities.

What context length does it support?

The model was trained with a context length of 65536 tokens and can use up to 64k tokens.

Other GGUFs of airoboros-l2-13b-2.1-YaRN-64k

Similar models

Browse all →

More from TheBloke

All by TheBloke →

More Airoboros models

All Airoboros →

Other medium models

All medium models →