3 ms·
This isn’t a fine tune. It’s not even a transformer like Qwen, Llama, GPT, Gemini, or Mistral. It’s another approach to language models based on RNNs
by jasonjmcghee 3y ago
This isn’t a fine tune. It’s not even a transformer like Qwen, Llama, GPT, Gemini, or Mistral. It’s another approach to language models based on RNNs
- Solvency 3y agoDoesn't the point remain though -- when/where do we demo this?
- junipertea 3y agoThere is Mamba 7B pre trained model on hugging face/ LM studio right now.