What are the hardware requirements for H2ogpt Oasst1 512 20B?

H2ogpt Oasst1 512 20B requires approximately 41.2 GB of VRAM and supports a context window of 2K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed H2ogpt Oasst1 512 20B and how large is it?

H2ogpt Oasst1 512 20B is developed by h2oai, a model with 20b parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does H2ogpt Oasst1 512 20B perform on standard benchmarks?

H2ogpt Oasst1 512 20B has the following published scores: MMLU 26.25. Compare against reference models on this page or on the LLM Explorer leaderboards.

H2ogpt Oasst1 512 20B by h2oai — VRAM 41.2GB, 2K context

Name: H2ogpt Oasst1 512 20B
Author: h2oai

H2ogpt Oasst1 512 20B is an open-source language model by h2oai. Features: 20b LLM, VRAM: 41.2GB, Context: 2K, License: apache-2.0, LLM Explorer Score: 0.11, Arc: 46.9, HellaSwag: 72.8, MMLU: 26.3, GSM8K: 3.2.

Dataset:h2oai/openassistant oa... Dataset:h2oai/openassistant oa... En Gpt Gpt neox Open-source Pytorch Region:us Sharded

Model Card on HF 🤗: https://huggingface.co/h2oai/h2ogpt-oasst1-512-20b

H2ogpt Oasst1 512 20B Parameters and Internals

Model Type

text generation, instruction-following, large language model

Use Cases

Primary Use Cases:

Instruction-following, Text generation

Limitations:

May produce biased, racist, offensive, or inappropriate content, Can produce incorrect, nonsensical, or irrelevant responses

Considerations:

Critical evaluation of the generated content is required. Use responsibly and ethically.

Additional Notes

This model is for commercial use and is optimized for use with the transformers library on machines with GPUs.

Supported Languages

en (English)

Training Details

Data Sources:

h2oai/openassistant_oasst1, h2oai/openassistant_oasst1_h2ogpt

Methodology:

fine-tuning

Model Architecture:

GPTNeoXForCausalLM

LLM Name	H2ogpt Oasst1 512 20B
Repository 🤗	https://huggingface.co/h2oai/h2ogpt-oasst1-512-20b
Model Size	20b
Required VRAM	41.2 GB
Updated	2026-05-18
Maintainer	h2oai
Model Type	gpt_neox
Model Files	4.9 GB: 1-of-9 4.9 GB: 2-of-9 4.8 GB: 3-of-9 4.8 GB: 4-of-9 4.9 GB: 5-of-9 4.8 GB: 6-of-9 4.8 GB: 7-of-9 4.9 GB: 8-of-9 2.4 GB: 9-of-9
Supported Languages	en
Model Architecture	GPTNeoXForCausalLM
License	apache-2.0
Context Length	2048
Model Max Length	2048
Transformers Version	4.28.1
Tokenizer Class	GPTNeoXTokenizer
Vocabulary Size	50432
Torch Data Type	float16

Rank the H2ogpt Oasst1 512 20B 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 H2ogpt Oasst1 512 20B

Best Alternatives	Context / RAM	Downloads	Likes
GPT Neox 20B	2K / 40.8 GB	794665	583
GPT NeoXT Chat Base 20B	2K / 41.2 GB	256	694
EleutherAI GPT Neox 20B 4bits	2K / 12.5 GB	8	0
...t Gm Oasst1 Multilang 1024 20B	2K / 41.2 GB	140	10
H2ogpt Gm Oasst1 En 1024 20B	2K / 41.2 GB	55	4
GPT Neox 20B Full Precision	2K / 82.5 GB	52	0
Oasst GPT Neox 20B 1000 Steps	2K / 41.2 GB	75	0
Oasst GPT Neox 20B 3000 Steps	2K / 41.2 GB	60	0
GPTNeoX 20B TestGen Dart V1.0	2K / 41.2 GB	5	2
GPT NeoX 20B Erebus	2K / 41.4 GB	576	88

Note: green Score (e.g. "73.2") means that the model is better than h2oai/h2ogpt-oasst1-512-20b.