LL

Meta-Llama-3-70B-Instruct-abliterated-v3.5

LlamaGGUF by failspyUpdated 2024-05-30
Download on Hugging Face
Parameters70.6B
Context8K
Layers80
Architecturellama
Smallest quant34.3 GB
Quick answer

Meta-Llama-3-70B-Instruct-abliterated-v3.5 needs about 61.1 GB of memory at its main file (a 57.9 GB file) with an 8K context window, which is more than a 48 GB GPU holds. This repository has 7 quants, from 34.3 GB to 141.1 GB.

Meta‑Llama‑3‑70B‑Instruct‑abliterated‑v3.5 – a 70 B instruct model with orthogonalized refusal directions

This 70‑billion‑parameter Llama model has been modified to reduce refusal behavior through orthogonalization. It retains the original instruction tuning and tokenizer. The model is available in several GGUF shards with varying file sizes.

Good at

  • Reduced refusal behavior
  • Large context support (8192 tokens)
  • Uncensored instruction responses

Limits

  • Requires high VRAM – 37.5 GB at 8 k context for smallest shard
  • No vision or multimodal capabilities
  • License is llama3, no commercial use without permission

Best for

  • Research on refusal mitigation
  • Developing instruction‑based chatbots
  • Testing large‑scale language models

Who it’s for. Researchers and developers needing a 70 B Llama model with minimized refusal behavior. Suitable for environments where uncensored instruction responses are required and sufficient GPU memory is available.

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
—34.3 GB 37.5 GB
Too big Get file →
—42.5 GB 45.7 GB
Too big Get file →
—49.9 GB 53.1 GB
Too big Get file →
—57.9 GB (3 parts) 61.1 GB
Too big Get file →
—75.0 GB (3 parts) 78.2 GB
Too big Get file →
—141.1 GB (6 parts) 144.3 GB
Too big Get file →
—141.1 GB (6 parts) 144.3 GB
Too big Get file →

The smallest shard is 34.27 GB and needs about 37.5 GB of VRAM at 8 k context. For 8 GB VRAM use is not feasible. If you have 16 GB VRAM consider the 42.52 GB shard, which requires ~45.7 GB VRAM. Larger VRAM (32 GB+) is needed for the 57.89 GB shard and above.

What GPU do you need for Meta-Llama-3-70B-Instruct-abliterated-v3.5?

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 the main file (42.5 GB file, about 45.7 GB in use).

Which Meta-Llama-3-70B-Instruct-abliterated-v3.5 quant should you choose?

If memory is tight, the smallest file is the first one in the table at 34.3 GB, with the quality trade-offs that come with fewer bits.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturellama
Parameters70.6B
Layers80
Trained context8,192 tokens
KV cache (fp16)320 KB per token

Source

Quantized byfailspy
FamilyLlama
Licensellama3
Downloads (HF)637
Last updated2024-05-30

Frequently asked questions about Meta-Llama-3-70B-Instruct-abliterated-v3.5

How much VRAM does Meta-Llama-3-70B-Instruct-abliterated-v3.5 need?

About 61.1 GB at its main file with an 8K context window. That is the 57.9 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 Meta-Llama-3-70B-Instruct-abliterated-v3.5 on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 37.5 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Meta-Llama-3-70B-Instruct-abliterated-v3.5 GGUF?

It depends on the quant: file is 34.3 GB, file is 42.5 GB, file is 49.9 GB, file is 57.9 GB, file is 75.0 GB, file is 141.1 GB and 1 more.

What context length does Meta-Llama-3-70B-Instruct-abliterated-v3.5 support?

The GGUF header lists a trained context length of 8,192 tokens (8K). The KV cache grows with the context you use: about 2.7 GB at 8K and 2.7 GB at 8K.

Where can I download Meta-Llama-3-70B-Instruct-abliterated-v3.5 GGUF?

The files are on Hugging Face in the failspy/Meta-Llama-3-70B-Instruct-abliterated-v3.5-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 Meta-Llama-3-70B-Instruct-abliterated-v3.5 use?

Hugging Face lists the license as llama3. 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?

No. The smallest shard requires about 37.5 GB of VRAM at 8 k context, so 8 GB is insufficient.

Does this model support vision or multimodal inputs?

No, it is a purely text‑based Llama model with no vision capabilities.

What license applies to this model?

The model is licensed under llama3, which restricts commercial use without permission.

Similar models

Browse all →

More from failspy

All by failspy →

More Llama models

All Llama →

Other large models

All large models →