What are the hardware requirements for Galactica Ref?

Galactica Ref requires approximately 0.5 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 Galactica Ref and how large is it?

Galactica Ref is developed by mariatager, a model with 125m 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 Galactica Ref?

Galactica Ref is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Galactica Ref by mariatager — VRAM 0.5GB, 2K context

Name: Galactica Ref
Author: mariatager

Galactica Ref is an open-source language model by mariatager. Features: 125m LLM, VRAM: 0.5GB, Context: 2K, LLM Explorer Score: 0.12.

Arxiv:1910.09700 Endpoints compatible Opt Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/mariatager/galactica_ref

Galactica Ref Parameters and Internals

LLM Name	Galactica Ref
Repository 🤗	https://huggingface.co/mariatager/galactica_ref
Model Size	125m
Required VRAM	0.5 GB
Updated	2026-07-20
Maintainer	mariatager
Model Type	opt
Model Files	0.5 GB
Model Architecture	OPTForCausalLM
Context Length	2048
Model Max Length	2048
Transformers Version	4.41.2
Tokenizer Class	PreTrainedTokenizerFast
Padding Token	[PAD]
Vocabulary Size	50003
Torch Data Type	float32
Activation Function	gelu

Rank the Galactica Ref 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 Galactica Ref

Best Alternatives	Context / RAM	Downloads	Likes
Opt 125M	2K / 0.3 GB	9414291	252
...25M Gqa Ub 6 Best For KV Cache	2K / 0.5 GB	207	0
...125M Qcqa Ub 6 Best For Q Loss	2K / 0.5 GB	57	0
...5M Qcqa Ub 6 Best For KV Cache	2K / 0.5 GB	58	0
Galactica 125M Cot	2K / 0.5 GB	5	0
Galactica 125M DPO	2K / 0.5 GB	7	0
Galactica 125M DPO Pos	2K / 0.5 GB	5	0
BertQA	2K / 0.5 GB	5	0
BertQA	2K / 0.5 GB	5	0
Opt 125M Quantized Brevitas	2K / GB	6	0

Note: green Score (e.g. "73.2") means that the model is better than mariatager/galactica_ref.