What are the hardware requirements for RCC Ins Reconstruction?

RCC Ins Reconstruction requires approximately 6.1 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 RCC Ins Reconstruction and how large is it?

RCC Ins Reconstruction is developed by fcyp, a model with 1.4b 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 RCC Ins Reconstruction?

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

RCC Ins Reconstruction by fcyp — VRAM 6.1GB, 2K context

Name: RCC Ins Reconstruction
Author: fcyp

RCC Ins Reconstruction is an open-source language model by fcyp. Features: 1.4b LLM, VRAM: 6.1GB, Context: 2K, License: apache-2.0, LLM Explorer Score: 0.13.

Arxiv:2406.06110 Endpoints compatible Gpt neox Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/fcyp/RCC_Ins_Reconstruction

RCC Ins Reconstruction Parameters and Internals

Model Type

text generation

Use Cases

Primary Use Cases:

QA tasks

Limitations:

Performance may degrade for sequences longer than the fine-tuning dataset length

Additional Notes

Employs FlashAttention-2.

Training Details

Methodology:

Recurrent Context Compression (RCC)

Context Length:

2048

Model Architecture:

Encoder-decoder based; Transformer for the decoder

Input Output

Input Format:

text prompt

Accepted Modalities:

text

Output Format:

text

Performance Tips:

Add specific code to prevent running out of memory during model evaluation.

Release Notes

Version:

RCC-Pythia-1.4b

Date:

2024/6/10

Notes:

Released the RCC-Pythia-1.4b model and related paper

LLM Name	RCC Ins Reconstruction
Repository 🤗	https://huggingface.co/fcyp/RCC_Ins_Reconstruction
Model Size	1.4b
Required VRAM	6.1 GB
Updated	2026-08-07
Maintainer	fcyp
Model Type	gpt_neox
Model Files	5.0 GB: 1-of-2 1.1 GB: 2-of-2
Model Architecture	GPTNeoXForCausalLM
License	apache-2.0
Context Length	2048
Model Max Length	2048
Transformers Version	4.40.2
Vocabulary Size	50304
Torch Data Type	bfloat16

Rank the RCC Ins Reconstruction 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 RCC Ins Reconstruction

Best Alternatives	Context / RAM	Downloads	Likes
Pythia 1.4B Deduped 8K Base	8K / 7.3 GB	4	1
Violet 1b4 Chat	4K / 2.8 GB	13	7
Pythia 1.4B Deduped 4K Base	4K / 6.1 GB	8	2
Statichh Pythia 1.4B Sft Bf16	2K / 5.7 GB	5	0
Pythia 1.4B	2K / 2.9 GB	95247	27
Pythia 1.4B Sft Full	2K / 2.8 GB	29	1
Pythia 1.4B Deduped	2K / 5.7 GB	14737	22
Pythia Delphi Suboptimal	2K / 5.7 GB	5	0
Pythia 1.4B Deduped Sharegpt	2K / 2.8 GB	129	2
Pythia 1.4B Deduped Sharegpt	2K / 2.8 GB	64	0

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