What are the hardware requirements for Ct2fast RedPajama INCITE Instruct 3B V1?

Ct2fast RedPajama INCITE Instruct 3B V1 requires approximately 5.5 GB of VRAM. Quantized variants may run on less VRAM; see the Quantized Models section on this page.

Who developed Ct2fast RedPajama INCITE Instruct 3B V1 and how large is it?

Ct2fast RedPajama INCITE Instruct 3B V1 is developed by michaelfeil, a model with 3b 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 Ct2fast RedPajama INCITE Instruct 3B V1?

Ct2fast RedPajama INCITE Instruct 3B V1 is hosted on Hugging Face and linked from this page. LLM Explorer also lists quantized variants and similar alternatives if available.

Ct2fast RedPajama INCITE Instruct 3B V1 by michaelfeil — VRAM 5.5GB

Name: Ct2fast RedPajama INCITE Instruct 3B V1
Author: michaelfeil

Ct2fast RedPajama INCITE Instruct 3B V1 is an open-source language model by michaelfeil. Features: 3b LLM, VRAM: 5.5GB, License: apache-2.0, Instruction-Based, LLM Explorer Score: 0.07.

Ctranslate2 Dataset:muennighoff/natural-in... Dataset:muennighoff/p3 Dataset:togethercomputer/redpa... En Endpoints compatible Float16 Instruct Int8 Region:us

Model Card on HF 🤗: https://huggingface.co/michaelfeil/ct2fast-RedPajama-INCITE-Instruct-3B-v1

Ct2fast RedPajama INCITE Instruct 3B V1 Parameters and Internals

Model Type

Language Model

Additional Notes

This model is a quantized version for 2x-4x speedup in inference using int8. This is intended to have identical license conditions as the original model.

Supported Languages

en (high)

Training Details

Data Sources:

togethercomputer/RedPajama-Data-1T, Muennighoff/P3, Muennighoff/natural-instructions

Data Volume:

131M tokens

Methodology:

fine-tuned for few-shot applications

Hardware Used:

8 A100

LLM Name	Ct2fast RedPajama INCITE Instruct 3B V1
Repository 🤗	https://huggingface.co/michaelfeil/ct2fast-RedPajama-INCITE-Instruct-3B-v1
Model Size	3b
Required VRAM	5.5 GB
Updated	2026-05-25
Maintainer	michaelfeil
Instruction-Based	Yes
Model Files	5.5 GB
Supported Languages	en
Model Architecture	AutoModel
License	apache-2.0
Model Max Length	2048
Tokenizer Class	GPTNeoXTokenizer

Rank the Ct2fast RedPajama INCITE Instruct 3B V1 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 Ct2fast RedPajama INCITE Instruct 3B V1

Best Alternatives	Context / RAM	Downloads	Likes
DAIsy Qwen Adapters	0K / 0.1 GB	10	0
CYPHER GUARD 3B	0K / 0.1 GB	14	1
Aether 2.3	0K / 0.1 GB	23	2
Testsentry Llama 3B	0K / 0 GB	33	0
...b33 8cc1 5be511e9bce8 5BossEnv	0K / 0.2 GB	69	0
Falcon 3B Finetune V1.0	0K / 0.1 GB	23	3
Falcon 3B Finetune V1.0	0K / 0.1 GB	4	4
Falcon 3B Finetune V1.0	0K / 0.1 GB	4	2
Falcon 3 3B Rp Lora Model	0K / 0.1 GB	0	2
Phi 4 Sky Lora	0K / 0.3 GB	0	2

Note: green Score (e.g. "73.2") means that the model is better than michaelfeil/ct2fast-RedPajama-INCITE-Instruct-3B-v1.