RWKV7-G1i-7.2B-20260805

This is RWKV-7 (G1i) model under flash-linear attention format.

Model introduction

RWKV-7 is an attention-free recurrent architecture with a constant-size recurrent state and constant inference work per generated token. Training remains parallelizable.

This checkpoint is a base model pretrained with web, code, synthetic, instruction, chat, and reasoning data. It is suitable for evaluation, post-training, and fine-tuning; the included chat template is a prompt interface, not a claim that the checkpoint is a safety-aligned assistant.

Model Details

• Developed by: Bo Peng, Yu Zhang, Songlin Yang, Ruichong Zhang, Zhiyuan Li
• Funded by: RWKV Project (Under LF AI & Data Foundation)
• Model type: RWKV7
• Language(s) (NLP): Multilingual
• License: Apache-2.0
• Parameter count: 7.2B
• Tokenizer: RWKV World tokenizer
• Vocabulary size: 65,536

Model Sources

• Repository: https://github.com/fla-org/flash-linear-attention ; https://github.com/BlinkDL/RWKV-LM
• Paper: https://arxiv.org/abs/2503.14456
• Model: rwkv7-g1i-7.2b-20260805-ctx16384.pth

Uses

Install flash-linear-attention and the latest version of transformers before using this model:

pip install git+https://github.com/fla-org/flash-linear-attention
pip install 'transformers>=4.48.0'

Direct Use

import fla
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('RWKV7-G1i-7.2B-20260805')
tokenizer = AutoTokenizer.from_pretrained('RWKV7-G1i-7.2B-20260805', trust_remote_code=True)
model = model.cuda()