RWKV is an RNN with transformer-level LLM performance. It can be directly trained like a GPT (parallelizable). So it's combining the best of RNN and transformer - great performance, fast inference, saves VRAM, fast training, "infinite" ctx_len, and free sentence embedding.
Apache License 2.0
12.32k
stars
838
forks
source link
Is it Feasible to Use ChatterBot as Another Model for Generating Optimal Initial States? #103
I came across your idea and was intrigued by it. I was wondering if it would be feasible to use ChatterBot as the another model for this project? Although I am not well-versed in this field, I am hopeful that someone with the necessary expertise can help bring this idea to fruition.
As for the idea itself, I find it very interesting. The ability to generate optimal initial states for any user input would be a valuable asset in many applications. I look forward to hearing more about your progress in developing this concept.
Dear BlinkDL,
I came across your idea and was intrigued by it. I was wondering if it would be feasible to use ChatterBot as the another model for this project? Although I am not well-versed in this field, I am hopeful that someone with the necessary expertise can help bring this idea to fruition.
As for the idea itself, I find it very interesting. The ability to generate optimal initial states for any user input would be a valuable asset in many applications. I look forward to hearing more about your progress in developing this concept.
Thank you for your time and consideration.
Sincerely, [win100]