FA

OpenAssistant falcon 40b megacode2 oasst

FalconGGUF by RichardErkhovUpdated 2024-09-02
Download on Hugging Face
Parameters41.3B
Context2K
Layers—
Architecturefalcon
Smallest quant9.9 GB
Quick answer

OpenAssistant falcon 40b megacode2 oasst needs about 15.4 GB of memory at Q4_K_M (a 13.3 GB file) with an 8K context window, so it fits on a 16 GB GPU. This repository has 15 quants, from 9.9 GB (IQ4_NL) to 26.1 GB (Q4_1).

OpenAssistant falcon 40b megacode2 oasst is a 41.3B-parameter Falcon model in GGUF format, quantized by RichardErkhov. Its trained context length is 2K tokens. The repository has 15 quant files, from 9.9 GB (IQ4_NL) to 26.1 GB (Q4_1).

Quants & memory

QuantFileEst. memoryvs VRAMResultDownload
IQ4_NL9.9 GB 10.7 GB
Too big Get file →
Q4_K_M13.3 GB 14.2 GB
Too big Get file →
Q2_K15.6 GB 16.6 GB
Too big Get file →
IQ3_S17.0 GB 18.1 GB
Too big Get file →
IQ3_XS17.5 GB 18.6 GB
Too big Get file →
Q3_K_S18.1 GB 19.1 GB
Too big Get file →
IQ3_M18.9 GB 20.0 GB
Too big Get file →
Q4_K_S19.6 GB 20.7 GB
Too big Get file →
Q4_K19.8 GB 20.9 GB
Too big Get file →
Q3_K19.9 GB 21.0 GB
Too big Get file →
Q3_K_M19.9 GB 21.0 GB
Too big Get file →
Q3_K_L21.4 GB 22.5 GB
Too big Get file →
IQ4_XS22.5 GB 23.7 GB
Too big Get file →
Q4_023.5 GB 24.7 GB
Too big Get file →
Q4_126.1 GB 27.3 GB
Too big Get file →

What GPU do you need for OpenAssistant falcon 40b megacode2 oasst?

These figures add the file size, an fp16 KV cache at an 8K context window 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: the best quant that fits is IQ4_NL (9.9 GB file, about 11.6 GB in use).
  • 16 GB VRAM: the best quant that fits is Q4_K_M (13.3 GB file, about 15.4 GB in use).
  • 24 GB VRAM: the best quant that fits is Q3_K (19.9 GB file, about 22.8 GB in use).
  • 32 GB VRAM: the best quant that fits is Q4_1 (26.1 GB file, about 29.7 GB in use).

Which OpenAssistant falcon 40b megacode2 oasst quant should you choose?

Q4_K_M is the usual starting point for OpenAssistant falcon 40b megacode2 oasst: it balances file size and quality well. Here it is a 13.3 GB file.

If memory is tight, the smallest file is IQ4_NL at 9.9 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.

New to quant names? Read GGUF quantization explained.

Specifications

Architecture from the GGUF header

Architecturefalcon
Parameters41.3B
Trained context2,048 tokens

Source

Quantized byRichardErkhov
FamilyFalcon
Downloads (HF)46
Last updated2024-09-02

Frequently asked questions about OpenAssistant falcon 40b megacode2 oasst

How much VRAM does OpenAssistant falcon 40b megacode2 oasst need?

About 15.4 GB at Q4_K_M with an 8K context window. That is the 13.3 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 OpenAssistant falcon 40b megacode2 oasst on an 8 GB GPU?

Not at an 8K context. The smallest file needs about 11.6 GB, so you would need a larger GPU or CPU offloading.

What is the file size of OpenAssistant falcon 40b megacode2 oasst GGUF?

It depends on the quant: IQ4_NL is 9.9 GB, Q4_K_M is 13.3 GB, Q2_K is 15.6 GB, IQ3_S is 17.0 GB, IQ3_XS is 17.5 GB, Q3_K_S is 18.1 GB and 9 more.

What context length does OpenAssistant falcon 40b megacode2 oasst support?

The GGUF header lists a trained context length of 2,048 tokens (2K).

Where can I download OpenAssistant falcon 40b megacode2 oasst GGUF?

The files are on Hugging Face in the RichardErkhov/OpenAssistant_-_falcon-40b-megacode2-oasst-gguf repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.

Similar models

Browse all →

More from RichardErkhov

All by RichardErkhov →

More Falcon models

All Falcon →

Other large models

All large models →