Downloads · 30 days
214
33% of all-time downloads
rwkv-2-pile-430m
BlinkDL/rwkv-2-pile-430m
rwkv-2-pile-430m is a text generation model from BlinkDL. Use it when you need the model to write or continue text. It is set up for rwkv. The card lists the license as apache-2.0.
RWKV-2 430M is a L24-D1024 causal language model trained on the Pile. See https://github.com/BlinkDL/RWKV-LM for details.
All-time downloads
654
Public
Repo size
1.7 GB
Likes
5
Public
Repo makeup
Click a slice to open those files.
.pth1.7 GB · 100%
At a glance
- Task
- Text Generation
- Library
- rwkv
- License
- apache-2.0
- Access
- Public
- Created
- Jun 23, 2022
- Updated
- Sep 14, 2022
- SHA
- 906332b9
Try a prompt
Model card
From the Hugging Face model README
RWKV-2 430M
Model Description
RWKV-2 430M is a L24-D1024 causal language model trained on the Pile. See https://github.com/BlinkDL/RWKV-LM for details.
At this moment you have to use my Github code (https://github.com/BlinkDL/RWKV-v2-RNN-Pile) to run it.
ctx_len = 768 n_layer = 24 n_embd = 1024
Final checkpoint: 20220615-10803.pth : Trained on the Pile for 331B tokens.
- Pile loss 2.349
- LAMBADA ppl 15.34, acc 42.42%
- PIQA acc 67.03%
- SC2016 acc 62.05%
- Hellaswag acc_norm 38.47%
- Task
- Text Generation
- Library
- rwkv
- License
- apache-2.0
- Languages
- en
- Created
- Jun 23, 2022
- Updated
- Sep 14, 2022