What are the hardware requirements for Math Custom Data?

Math Custom Data requires approximately 15.2 GB of VRAM and supports a context window of 128K tokens. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Math Custom Data and how large is it?

Math Custom Data is developed by amphora, a model with 7.6b 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 Math Custom Data?

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

Math Custom Data by amphora — VRAM 15.2GB, 128K context

Name: Math Custom Data
Author: amphora

Math Custom Data is an open-source language model by amphora. Features: 7.6b LLM, VRAM: 15.2GB, Context: 128K, LLM Explorer Score: 0.23.

Arxiv:1910.09700 Conversational Endpoints compatible Qwen2 Region:us Safetensors

Model Card on HF 🤗: https://huggingface.co/amphora/math-custom-data

Math Custom Data Parameters and Internals

LLM Name	Math Custom Data
Repository 🤗	https://huggingface.co/amphora/math-custom-data
Model Size	7.6b
Required VRAM	15.2 GB
Updated	2026-07-26
Maintainer	amphora
Model Type	qwen2
Model Files	15.2 GB
Model Architecture	Qwen2ForCausalLM
Context Length	131072
Model Max Length	131072
Transformers Version	5.2.0
Tokenizer Class	Qwen2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	152064
Errors	replace

Rank the Math Custom Data 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 Math Custom Data

Best Alternatives	Context / RAM	Downloads	Likes
Quasar 3.7	986K / 15.2 GB	805	0
Quasar 3.3 Max	986K / 15.2 GB	132	0
MQ Coldbrew Base	128K / 15.2 GB	761	0
MawaredT1	128K / 15.2 GB	2318	1
Mergekit 01	128K / 16.2 GB	6	0
BBALAW1.6	128K / 15.2 GB	7	0
BBAI 525 Tsu GZ Xia0	128K / 15.2 GB	8	0
Nature Reason 1.2 Reallysmall	128K / 15.2 GB	27	1
Exp 3 Q R	128K / 15.2 GB	6	0
Marco LLM AR	128K / 15.3 GB	1626	0

Note: green Score (e.g. "73.2") means that the model is better than amphora/math-custom-data.