What are the hardware requirements for Saily 220B GPTQ?

Saily 220B GPTQ requires approximately 105.2 GB of VRAM and supports a context window of 4K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Saily 220B GPTQ and how large is it?

Saily 220B GPTQ is developed by TheBloke, a model with 220b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

Where can I download or evaluate Saily 220B GPTQ?

Saily 220B GPTQ is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Saily 220B GPTQ by TheBloke — VRAM 105.2GB, 4K context

Name: Saily 220B GPTQ
Author: TheBloke

Saily 220B GPTQ is an open-source language model by TheBloke. Features: 220b LLM, VRAM: 105.2GB, Context: 4K, License: llama2, Quantized, LLM Explorer Score: 0.1.

4-bit Base model:deepnight-research/... Base model:quantized:deepnight... Dataset:eleutherai/pile Dataset:meta-math/metamathqa Dataset:tiiuae/falcon-refinedw... En Gptq Llama Quantized Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/TheBloke/Saily_220B-GPTQ

Saily 220B GPTQ Parameters and Internals

Model Type

llama

Additional Notes

Limitations include the possibility of incorrect or biased content generation.

Supported Languages

English (primary)

Training Details

Data Sources:

tiiuae/falcon-refinedweb, EleutherAI/pile, meta-math/MetaMathQA, Unnatural Code

Methodology:

Built on top of Llama2-70B merges, 10 fine-tuned models; multiple merges created keeping Logical-Understanding and Reasoning models constant.

Hardware Used:

4 x A100 80GB GPUs, 2 x A100 80GB GPUs

Model Architecture:

Llama2-70B fine-tuned models with multiple merges.

Input Output

Input Format:

Use the Alpaca Prompt Format with Instruction and Response tags.

Output Format:

Textual response output following given instruction.

Performance Tips:

Load in 4bit or 8bit for optimization; please use INSTRUCT or CHAT-INSTRUCT mode in Text-Generation-WebUI.

Release Notes

Version:

Date:

17th December, 2023

Notes:

Released as a powerful model built on Llama2-70B merges; 10 models fine-tuned on specific domains; uses public datasets and internal transcriptions.

LLM Name	Saily 220B GPTQ
Repository 🤗	https://huggingface.co/TheBloke/Saily_220B-GPTQ
Model Name	Saily 220B
Model Creator	DEEPNIGHT
Base Model(s)	Saily 220B deepnight-research/Saily_220B
Model Size	220b
Required VRAM	105.2 GB
Updated	2026-08-02
Maintainer	TheBloke
Model Type	llama
Model Files	49.0 GB: 1-of-3 48.9 GB: 2-of-3 7.3 GB: 3-of-3
Supported Languages	en
GPTQ Quantization	Yes
Quantization Type	gptq
Model Architecture	LlamaForCausalLM
License	llama2
Context Length	4096
Model Max Length	4096
Transformers Version	4.37.0.dev0
Tokenizer Class	LlamaTokenizer
Vocabulary Size	32000
Torch Data Type	float16

Rank the Saily 220B GPTQ capabilities

Have you tried this model? Rate its performance — this feedback helps the ML community find the right model for their needs.

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

Best Alternatives to Saily 220B GPTQ

Best Alternatives	Context / RAM	Downloads	Likes
...l Llama 220M GQA 32K Theta Sft	32K / 0.4 GB	6	2
Smol Llama 220M GQA 32K Theta	32K / 0.4 GB	8	1
Nuevamol 220M Base	8K / 0.4 GB	6	0
Saily 220B AWQ	4K / 109.1 GB	5	0
Saily 220B	4K / 417 GB	96	20
Smol Llama 220M GQA	2K / 0.4 GB	326	13
Smol Llama 220M Openhermes	2K / 0.4 GB	113	5
Smol Llama 220M Open Instruct	2K / 0.4 GB	6	2
...mol Llama 220M GQA Fineweb Edu	2K / 0.4 GB	14	1
Smol Llama 220M Bees Internal	2K / 0.4 GB	6	1

Note: green Score (e.g. "73.2") means that the model is better than TheBloke/Saily_220B-GPTQ.