What are the hardware requirements for Llama2 22B Daydreamer V3?

Llama2 22B Daydreamer V3 requires approximately 43.7 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 Llama2 22B Daydreamer V3 and how large is it?

Llama2 22B Daydreamer V3 is developed by nkpz, a model with 22b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does Llama2 22B Daydreamer V3 perform on standard benchmarks?

Llama2 22B Daydreamer V3 has the following published scores: MMLU 52.49. Compare against reference models on this page or on the LLM Explorer leaderboards.

Llama2 22B Daydreamer V3 by nkpz — VRAM 43.7GB, 4K context

Name: Llama2 22B Daydreamer V3
Author: nkpz

Llama2 22B Daydreamer V3 is an open-source language model by nkpz. Features: 22b LLM, VRAM: 43.7GB, Context: 4K, License: other, LLM Explorer Score: 0.11, Arc: 56.1, HellaSwag: 80.1, MMLU: 52.5, GSM8K: 3.8.

Endpoints compatible Llama Pytorch Region:us Sharded

Model Card on HF 🤗: https://huggingface.co/nkpz/llama2-22b-daydreamer-v3

Llama2 22B Daydreamer V3 Parameters and Internals

Model Type

Roleplay, Q&A, Instruction Following, Conversational

Use Cases

Areas:

Roleplay, Instruction following, Conversational applications

Additional Notes

Model has been enhanced to roleplay and imitate speech styles effectively. Exhibits a dry and accurate response style for informational queries but is also capable of delivering playful responses for casual prompts.

Supported Languages

English (Proficient)

Training Details

Data Sources:

Instruction data, Roleplay data, Conversational data, Jondurbin Airoboros GPT4 1.4.1 MPT Dataset excluding coding rows

Hardware Used:

NVIDIA 3090

Input Output

Accepted Modalities:

Text

LLM Name	Llama2 22B Daydreamer V3
Repository 🤗	https://huggingface.co/nkpz/llama2-22b-daydreamer-v3
Model Size	22b
Required VRAM	43.7 GB
Updated	2026-07-31
Maintainer	nkpz
Model Type	llama
Model Files	9.8 GB: 1-of-5 10.0 GB: 2-of-5 9.9 GB: 3-of-5 9.9 GB: 4-of-5 4.1 GB: 5-of-5
Model Architecture	LlamaForCausalLM
License	other
Context Length	4098
Model Max Length	4098
Transformers Version	4.31.0
Tokenizer Class	LlamaTokenizer
Beginning of Sentence Token	<s>
End of Sentence Token	</s>
Unk Token	<unk>
Vocabulary Size	32000
Torch Data Type	float16

Rank the Llama2 22B Daydreamer V3 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

Quantized Models of the Llama2 22B Daydreamer V3

Model	Likes	Downloads	VRAM
Llama2 22B Daydreamer V3 GPT	0	7	12 GB
Llama2 22B Daydreamer V3 GGUF	2	362	9 GB
Llama2 22B Daydreamer V3 AWQ	2	4	12 GB
Llama2 22B Daydreamer V3 GPTQ	6	10	12 GB
Llama2 22B Daydreamer V3 GGML	6	5	9 GB

Best Alternatives to Llama2 22B Daydreamer V3

Best Alternatives	Context / RAM	Downloads	Likes
EuroLLM 22B Instruct 2512	32K / 45.5 GB	3342	63
EuroLLM 22B 2512	32K / 45.5 GB	2766	19
Calm3 22B Chat	16K / 44.9 GB	878	85
Calm3 22B RP V2	16K / 44.9 GB	21	20
Calm3 22B RP V0.1	16K / 44.9 GB	10	0
Yousei 22B	4K / 44.5 GB	97	2
Platypus 2 22B Relora	4K / 43.7 GB	93	1
Llama2 22B Blocktriangular	4K / 43.7 GB	368	4
Llama2 22B	4K / 43.7 GB	109	46
Llama2 22B Daydreamer V2	4K / 43.7 GB	7	2

Note: green Score (e.g. "73.2") means that the model is better than nkpz/llama2-22b-daydreamer-v3.