SW

Swallow 70B instruct

SwallowGGUF by TheBlokeUpdated 2023-12-19
Download on Hugging Face
Parameters69.2B
Context4K
Layers80
Architecturellama
Smallest quant29.4 GB
Quick answer

Swallow 70B instruct needs about 44.7 GB of memory at Q4_K_M (a 41.5 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 10 quants, from 29.4 GB (Q2_K) to 48.9 GB (Q5_K_M).

Swallow 70B Instruct – a large Japanese‑English instruction‑tuned LLaMA‑2 model

Swallow 70B Instruct is a 69.2 billion‑parameter LLaMA‑2 based model tuned with supervised fine‑tuning. It supports Japanese and English and is available in GGUF format for llama.cpp and compatible clients. The model was released on April 26, 2024 as part of the Swallow‑70b‑instruct‑v0.1 preview.

Good at

  • Japanese language tasks
  • Instruction following
  • Large‑context reasoning

Limits

  • No vision support
  • Requires 8 k context for memory estimates
  • License is llama2

Best for

  • Japanese‑English translation
  • Long‑form content generation
  • Chatbot development

Who it’s for. Developers and researchers needing a high‑capacity instruction model that handles both Japanese and English. The model is suitable for local deployment via llama.cpp or compatible GUIs.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K29.4 GB 31.2 GB
Too big Get file →
Q3_K_S30.0 GB 31.9 GB
Too big Get file →
Q3_K_M33.3 GB 35.1 GB
Too big Get file →
Q3_K_L36.3 GB 38.1 GB
Too big Get file →
Q4_039.0 GB 40.8 GB
Too big Get file →
Q4_K_S39.2 GB 41.0 GB
Too big Get file →
Q4_K_M41.5 GB 43.4 GB
Too big Get file →
Q5_047.6 GB 49.4 GB
Too big Get file →
Q5_K_S47.6 GB 49.4 GB
Too big Get file →
Q5_K_M48.9 GB 50.7 GB
Too big Get file →

For 8 GB VRAM use Q4_0 (file size 39 GB, ~42.2 GB memory at 8 k context). For 16 GB VRAM use Q5_0 (file size 47.6 GB, ~50.8 GB memory). For 32 GB VRAM use Q5_K_M (file size 48.89 GB, ~52.1 GB memory). Estimates are approximate.

What GPU do you need for Swallow 70B 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 Q4_K_M (41.5 GB file, about 44.7 GB in use).

Which Swallow 70B instruct quant should you choose?

Q4_K_M is the usual starting point for Swallow 70B instruct: it balances file size and quality well. Here it is a 41.5 GB file.

If memory is tight, the smallest file is Q2_K at 29.4 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
Parameters69.2B
Layers80
Trained context4,096 tokens
KV cache (fp16)320 KB per token

Source

Quantized byTheBloke
FamilySwallow
Licensellama2
Downloads (HF)2,646
Last updated2023-12-19

Frequently asked questions about Swallow 70B instruct

How much VRAM does Swallow 70B instruct need?

About 44.7 GB at Q4_K_M with an 8K context window. That is the 41.5 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 Swallow 70B instruct on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 32.6 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Swallow 70B instruct GGUF?

It depends on the quant: Q2_K is 29.4 GB, Q3_K_S is 30.0 GB, Q3_K_M is 33.3 GB, Q3_K_L is 36.3 GB, Q4_0 is 39.0 GB, Q4_K_S is 39.2 GB and 4 more.

What context length does Swallow 70B instruct 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 Swallow 70B instruct GGUF?

The files are on Hugging Face in the TheBloke/Swallow-70B-instruct-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 Swallow 70B instruct 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 Swallow 70B Instruct on 8 GB VRAM?

The Q4_0 quant requires about 42.2 GB memory at 8 k context, so 8 GB VRAM is insufficient.

Does Swallow 70B Instruct support vision?

No, the model has no vision capabilities.

What license applies to Swallow 70B Instruct?

The model is licensed under llama2.

Other GGUFs of Swallow-70b-instruct-hf

Similar models

Browse all →

More from TheBloke

All by TheBloke →

More Swallow models

All Swallow →

Other large models

All large models →