TR-3B
TR-3B
> hey looking at the merged PR should this be a open issue be closed ? yeah, had the same question, is this an issue now?
This is my implementation with llama 3.1 8b model, which pretty much solves issue #1400, This is the reference notebook @danielhanchen for that issue.. [llama 3.1 tool calling](https://colab.research.google.com/gist/MagellaX/2dc7c6b4faf7ae49f17eac9945bacc7c/tool-calling.ipynb)
[save.zip](https://github.com/user-attachments/files/20012996/save.zip) This might work, but it's an enhanced version, a more bullet-proof implementation
@alckasoc any thoughts? u can merge this
> Hi @MagellaX , sorry for the late response! Thank you for the contribution! > > Some questions: > > * From what I understand, this PR is to enforce...
Reminder that this is a foundational LoRA support, meaning that from here we can bring things/more features to MLC-LLM such as multi-LoRA batching (pending upstream TVM/Relax changes), dynamic LoRA switching...
@junrushao @MasterJH5574 ANY TAKES?
@gemini-code-assist re-run the review
@hijkzzz any thoughts?
> This is a great MR. I’ll need some time to go through it carefully. The current implementation may have some issues with vLLM weight synchronization and HF weight saving...