JA

Jais Adapted 70b Chat

JaisGGUF by RichardErkhovUpdated 2024-08-10
Download on Hugging Face
Parameters69.5B
Context4K
Layers80
Architecturellama
Smallest quant25.8 GB
Quick answer

Jais Adapted 70b Chat needs about 45.0 GB of memory at Q4_K_M (a 41.8 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 22 quants, from 25.8 GB (Q2_K) to 73.9 GB (Q8_0).

Jais Adapted 70b Chat – 69.5B parameter bilingual English‑Arabic LLM

Jais Adapted 70b Chat is a 69.5B parameter, transformer‑based, decoder‑only model. It is instruction fine‑tuned for dialogue in Arabic and English. The model supports a 4096 token context window and is released under the Apache‑2.0 license.

Good at

  • Arabic language generation
  • English language generation
  • Instruction‑driven dialogue

Limits

  • No vision support
  • Large VRAM requirement for higher quantizations
  • No mixed‑precision inference on low‑end GPUs

Best for

  • Chatbots for Arabic and English
  • Bilingual content creation
  • Research in Arabic NLP

Who it’s for. Developers building bilingual chat applications or researchers needing a large Arabic‑English model. The model is suitable for environments with sufficient GPU memory to run the chosen quantization.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K25.8 GB 27.6 GB
Too big Get file →
IQ3_XS28.6 GB 30.5 GB
Too big Get file →
IQ3_S30.2 GB 32.1 GB
Too big Get file →
Q3_K_S30.2 GB 32.1 GB
Too big Get file →
IQ3_M31.3 GB 33.1 GB
Too big Get file →
Q3_K33.6 GB 35.4 GB
Too big Get file →
Q3_K_M33.6 GB 35.4 GB
Too big Get file →
Q3_K_L36.5 GB 38.3 GB
Too big Get file →
IQ4_XS37.6 GB 39.4 GB
Too big Get file →
Q4_039.2 GB 41.1 GB
Too big Get file →
IQ4_NL39.6 GB 41.5 GB
Too big Get file →
Q4_K_S39.6 GB 41.5 GB
Too big Get file →
Q4_K41.8 GB (2 parts) 43.6 GB
Too big Get file →
Q4_K_M41.8 GB (2 parts) 43.6 GB
Too big Get file →
Q4_143.5 GB (2 parts) 45.4 GB
Too big Get file →
Q5_047.9 GB (2 parts) 49.7 GB
Too big Get file →
Q5_K_S47.9 GB (2 parts) 49.7 GB
Too big Get file →
Q5_K49.1 GB (2 parts) 51.0 GB
Too big Get file →
Q5_K_M49.1 GB (2 parts) 51.0 GB
Too big Get file →
Q5_152.2 GB (2 parts) 54.0 GB
Too big Get file →
Q6_K57.0 GB (2 parts) 58.9 GB
Too big Get file →
Q8_073.9 GB (2 parts) 75.7 GB
Too big Get file →

For 8 GB VRAM, use Q2_K (25.76 GB file, ~28.9 GB memory). For 12 GB VRAM, Q3_K_S (30.25 GB file, ~33.4 GB memory). For 24 GB VRAM, Q4_K (41.79 GB file, ~45 GB memory). These estimates are approximate.

What GPU do you need for Jais Adapted 70b Chat?

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: the best quant that fits is IQ3_XS (28.6 GB file, about 31.8 GB in use).
  • 48 GB VRAM: the best quant that fits is Q4_1 (43.5 GB file, about 46.7 GB in use).

Which Jais Adapted 70b Chat quant should you choose?

Q4_K_M is the usual starting point for Jais Adapted 70b Chat: it balances file size and quality well. Here it is a 41.8 GB file.

The Q8_0 file is 73.9 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 25.8 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
Parameters69.5B
Layers80
Trained context4,096 tokens
KV cache (fp16)320 KB per token

Source

Quantized byRichardErkhov
FamilyJais
Licenseapache-2.0
Downloads (HF)2,550
Last updated2024-08-10

Frequently asked questions about Jais Adapted 70b Chat

How much VRAM does Jais Adapted 70b Chat need?

About 45.0 GB at Q4_K_M with an 8K context window. That is the 41.8 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 Jais Adapted 70b Chat on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 28.9 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Jais Adapted 70b Chat GGUF?

It depends on the quant: Q2_K is 25.8 GB, IQ3_XS is 28.6 GB, IQ3_S is 30.2 GB, Q3_K_S is 30.2 GB, IQ3_M is 31.3 GB, Q3_K is 33.6 GB and 16 more.

What context length does Jais Adapted 70b Chat 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 Jais Adapted 70b Chat GGUF?

The files are on Hugging Face in the RichardErkhov/inceptionai_-_jais-adapted-70b-chat-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 Jais Adapted 70b Chat use?

Hugging Face lists the license as apache-2.0. 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 Jais Adapted 70b Chat on 8 GB VRAM?

Use the Q2_K quantization; it requires about 28.9 GB of VRAM at 8k context, which exceeds 8 GB, so it is not feasible.

Does the model support vision input?

No, the model only accepts text input.

What license applies to this model?

The model is released under the Apache‑2.0 license.

Similar models

Browse all →

More from RichardErkhov

All by RichardErkhov →

More Jais models

All Jais →

Other large models

All large models →