What are the hardware requirements for TinyStories 33M?

TinyStories 33M requires approximately 0.3 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 TinyStories 33M and how large is it?

TinyStories 33M is developed by roneneldan, a model with 33m parameters. The model is published as open weights on Hugging Face and indexed on LLM Explorer with full benchmark history.

How does TinyStories 33M perform on standard benchmarks?

TinyStories 33M has the following published scores: MMLU 23.82. Compare against reference models on this page or on the LLM Explorer leaderboards.

TinyStories 33M by roneneldan — VRAM 0.3GB, 2K context

Name: TinyStories 33M
Author: roneneldan

TinyStories 33M is an open-source language model by roneneldan. Features: 33m LLM, VRAM: 0.3GB, Context: 2K, LLM Explorer Score: 0.15, Arc: 24.2, HellaSwag: 25.7, MMLU: 23.8.

Arxiv:2305.07759 Dataset:roneneldan/tinystories Endpoints compatible Gpt neo Pytorch Region:us

Model Card on HF 🤗: https://huggingface.co/roneneldan/TinyStories-33M

TinyStories 33M Parameters and Internals

Model Type

text generation

Training Details

Data Sources:

roneneldan/TinyStories

Context Length:

512

Model Architecture:

GPT-Neo

Input Output

Input Format:

tokenized text

Accepted Modalities:

text

Output Format:

generated text

LLM Name	TinyStories 33M
Repository 🤗	https://huggingface.co/roneneldan/TinyStories-33M
Model Size	33m
Required VRAM	0.3 GB
Updated	2026-07-10
Maintainer	roneneldan
Model Type	gpt_neo
Model Files	0.3 GB
Model Architecture	GPTNeoForCausalLM
Context Length	2048
Model Max Length	2048
Transformers Version	4.28.0
Tokenizer Class	GPT2Tokenizer
Beginning of Sentence Token	<\|endoftext\|>
End of Sentence Token	<\|endoftext\|>
Unk Token	<\|endoftext\|>
Vocabulary Size	50257
Torch Data Type	float32
Activation Function	gelu_new
Errors	replace

Rank the TinyStories 33M 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 TinyStories 33M

Best Alternatives	Context / RAM	Downloads	Likes
...eo 33M Simplewiki 2048 Scratch	2K / 0.3 GB	3	1
TinyStories 33M Ds	2K / GB	7	0
TinyStories Instruct 33M	2K / 0.3 GB	2599	10
...Tinystories 33M Epoch10 Merged	2K / 0.3 GB	15	2
TinyStories 33M Finetuned	2K / 0.3 GB	12	0
TinyStories 2Layers 33M	2K / 0.3 GB	510	5
...nyStories Instruct 2Layers 33M	2K / 0.3 GB	346	7