TO

tokyotech llm Swallow 7b instruct hf

TokyotechGGUF by RichardErkhovUpdated 2024-05-10
Download on Hugging Face
Parameters6.83B
Context4K
Layers32
Architecturellama
Smallest quant2.6 GB
Quick answer

tokyotech llm Swallow 7b instruct hf 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 21 quants, from 2.6 GB (Q2_K) to 5.6 GB (Q6_K).

Swallow 7b instruct HF – a 6.83B LLaMA‑based model tuned for Japanese and English

Swallow 7b instruct HF is a 6.83B parameter LLaMA model fine‑tuned with Japanese data. It supports instruction following and uses a broadened tokenizer for efficient Japanese representation. The model is available in multiple GGUF quantizations.

Good at

  • Instruction following
  • Japanese and English text generation
  • Efficient token usage

Limits

  • No vision support
  • Context limited to 4096 tokens
  • License unspecified

Best for

  • Chatbots needing Japanese support
  • Text summarization in bilingual contexts
  • Lightweight inference on modest GPUs

Who it’s for. Developers building bilingual LLM applications who need a compact model with Japanese proficiency. The model is suitable for edge deployment where GPU memory is limited.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K2.6 GB 5.2 GB
Fits Get file →
IQ3_XS2.9 GB 5.5 GB
Fits Get file →
IQ3_S3.0 GB 5.7 GB
Fits Get file →
Q3_K_S3.0 GB 5.7 GB
Fits Get file →
IQ3_M3.2 GB 5.8 GB
Fits Get file →
Q3_K3.4 GB 6.0 GB
Fits Get file →
Q3_K_M3.4 GB 6.0 GB
Fits Get file →
Q3_K_L3.7 GB 6.3 GB
Fits Get file →
IQ4_XS3.7 GB 6.4 GB
Fits Get file →
Q4_03.9 GB 6.5 GB
Fits Get file →
IQ4_NL3.9 GB 6.6 GB
Fits Get file →
Q4_K_S3.9 GB 6.6 GB
Fits Get file →
Q4_K4.1 GB 6.8 GB
Fits Get file →
Q4_K_M4.1 GB 6.8 GB
Fits Get file →
Q4_14.3 GB 7.0 GB
Fits Get file →
Q5_04.7 GB 7.4 GB
Tight Get file →
Q5_K_S4.7 GB 7.4 GB
Tight Get file →
Q5_K4.9 GB 7.5 GB
Tight Get file →
Q5_K_M4.9 GB 7.5 GB
Tight Get file →
Q5_15.1 GB 7.8 GB
Tight Get file →
Q6_K5.6 GB 8.3 GB
Too big Get file →

For 8 GB VRAM, use Q4_K or Q4_K_M (≈8.9 GB memory at 8k context). For 12 GB VRAM, Q5_K or Q5_K_M (≈9.6 GB). For 16 GB VRAM, Q6_K (≈10.4 GB). Memory estimates are approximate.

What GPU do you need for tokyotech llm Swallow 7b instruct hf?

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 IQ3_M (3.2 GB file, about 8.0 GB in use).
  • 12 GB VRAM: the best quant that fits is Q6_K (5.6 GB file, about 10.4 GB in use).

Which tokyotech llm Swallow 7b instruct hf quant should you choose?

Q4_K_M is the usual starting point for tokyotech llm Swallow 7b instruct hf: it balances file size and quality well. Here it is a 4.1 GB file.

If memory is tight, the smallest file is Q2_K at 2.6 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
Parameters6.83B
Layers32
Trained context4,096 tokens
KV cache (fp16)512 KB per token

Source

Quantized byRichardErkhov
FamilyTokyotech
Downloads (HF)2,383
Last updated2024-05-10

Frequently asked questions about tokyotech llm Swallow 7b instruct hf

How much VRAM does tokyotech llm Swallow 7b instruct hf 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 tokyotech llm Swallow 7b instruct hf on an 8 GB GPU?

Yes, with IQ3_M: the 3.2 GB file needs about 8.0 GB at 8K context. Larger quants in this repository do not fit in 8 GB.

What is the file size of tokyotech llm Swallow 7b instruct hf GGUF?

It depends on the quant: Q2_K is 2.6 GB, IQ3_XS is 2.9 GB, IQ3_S is 3.0 GB, Q3_K_S is 3.0 GB, IQ3_M is 3.2 GB, Q3_K is 3.4 GB and 15 more.

What context length does tokyotech llm Swallow 7b instruct hf support?

The GGUF header lists a trained context length of 4,096 tokens (4K). The KV cache grows with the context you use: about 4.3 GB at 8K and 2.1 GB at 4K.

Where can I download tokyotech llm Swallow 7b instruct hf GGUF?

The files are on Hugging Face in the RichardErkhov/tokyotech-llm_-_Swallow-7b-instruct-hf-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 Swallow 7b instruct HF on 8 GB VRAM?

Yes, using Q4_K or Q4_K_M quantizations will fit within 8 GB VRAM, with memory usage around 8.9 GB at 8k context.

Does the model support image inputs?

No, Swallow 7b instruct HF does not support vision inputs.

What languages can the model handle?

The model is trained for Japanese and English.

Similar models

Browse all →

More from RichardErkhov

All by RichardErkhov →

More Tokyotech models

All Tokyotech →

Other small models

All small models →