ME

Merged RP Stew V2 68B

MergedGGUF by mradermacherUpdated 2024-05-06
Download on Hugging Face
Parameters67.9B
Context195K
Layers120
Architecturellama
Smallest quant25.1 GB
Quick answer

Merged RP Stew V2 68B needs about 45.2 GB of memory at Q4_K_M (a 40.7 GB file) with an 8K context window, so it fits on a 48 GB GPU. This repository has 12 quants, from 25.1 GB (Q2_K) to 48.0 GB (Q5_K_M). Longer context needs more: at 32K tokens the KV cache alone adds about 16.1 GB.

Merged RP Stew V2 68B is a 67.9B-parameter Merged model in GGUF format, quantized by mradermacher. Its trained context length is 195K tokens. The repository has 12 quant files, from 25.1 GB (Q2_K) to 48.0 GB (Q5_K_M).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
Q2_K25.1 GB 29.6 GB
Too big Get file →
IQ3_XS27.9 GB 32.4 GB
Too big Get file →
Q3_K_S29.3 GB 33.9 GB
Too big Get file →
IQ3_S29.5 GB 34.0 GB
Too big Get file →
IQ3_M30.6 GB 35.1 GB
Too big Get file →
Q3_K_M32.8 GB 37.3 GB
Too big Get file →
Q3_K_L35.7 GB 40.2 GB
Too big Get file →
IQ4_XS36.7 GB 41.2 GB
Too big Get file →
Q4_K_S38.6 GB 43.1 GB
Too big Get file →
Q4_K_M40.7 GB 45.2 GB
Too big Get file →
Q5_K_S46.7 GB 51.2 GB
Too big Get file →
Q5_K_M48.0 GB 52.5 GB
Too big Get file →

What GPU do you need for Merged RP Stew V2 68B?

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 Q2_K (25.1 GB file, about 29.6 GB in use).
  • 48 GB VRAM: the best quant that fits is Q4_K_M (40.7 GB file, about 45.2 GB in use).

Which Merged RP Stew V2 68B quant should you choose?

Q4_K_M is the usual starting point for Merged RP Stew V2 68B: it balances file size and quality well. Here it is a 40.7 GB file.

If memory is tight, the smallest file is Q2_K at 25.1 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
Parameters67.9B
Layers120
Trained context200,000 tokens
KV cache (fp16)480 KB per token

Source

Quantized bymradermacher
FamilyMerged
Licenseother
Downloads (HF)115
Last updated2024-05-06

Frequently asked questions about Merged RP Stew V2 68B

How much VRAM does Merged RP Stew V2 68B need?

About 45.2 GB at Q4_K_M with an 8K context window. That is the 40.7 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 Merged RP Stew V2 68B on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 29.6 GB, so you would need a larger GPU or CPU offloading.

What is the file size of Merged RP Stew V2 68B GGUF?

It depends on the quant: Q2_K is 25.1 GB, IQ3_XS is 27.9 GB, Q3_K_S is 29.3 GB, IQ3_S is 29.5 GB, IQ3_M is 30.6 GB, Q3_K_M is 32.8 GB and 6 more.

What context length does Merged RP Stew V2 68B support?

The GGUF header lists a trained context length of 200,000 tokens (195K). The KV cache grows with the context you use: about 4.0 GB at 8K and 16.1 GB at 32K.

Where can I download Merged RP Stew V2 68B GGUF?

The files are on Hugging Face in the mradermacher/Merged-RP-Stew-V2-68B-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 Merged RP Stew V2 68B use?

Hugging Face lists the license as other. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.

Similar models

Browse all →

More from mradermacher

All by mradermacher →

More Merged models

All Merged →

Other large models

All large models →