GPTCache
GPTCache copied to clipboard
[Enhancement]: Add RWKV model support (RWKV is a 100% RNN Language Model - ctxlen 8192 models available, longer ctxlen soon)
What would you like to be added?
RWKV Raven 7B Gradio Demo: https://huggingface.co/spaces/BlinkDL/Raven-RWKV-7B
Use rwkv.cpp for CPU INT4 / INT8: https://github.com/saharNooby/rwkv.cpp
Github project: https://github.com/BlinkDL/ChatRWKV
Sample code using rwkv pip package: https://github.com/BlinkDL/ChatRWKV/blob/main/v2/benchmark_more.py
Please let me know if you have any questions :)
Why is this needed?
No response
Anything else?
No response
What would you like to be added?
RWKV Raven 7B Gradio Demo: https://huggingface.co/spaces/BlinkDL/Raven-RWKV-7B
Use rwkv.cpp for CPU INT4 / INT8: https://github.com/saharNooby/rwkv.cpp
Github project: https://github.com/BlinkDL/ChatRWKV
Sample code using rwkv pip package: https://github.com/BlinkDL/ChatRWKV/blob/main/v2/benchmark_more.py
Please let me know if you have any questions :)
Why is this needed?
No response
Anything else?
No response
@BlinkDL want to take the RWKV features? the RWKV has very long context window, which is great
We are adding the rwkv embedding, see: #363 If you have other ideas, please feel free to give a comment
If you have other ideas, please open a new issue