Ben

same issue, how to resolve that? thx

Do we really need the `LLAMA_TOKENIZER_PATH` as I can't see it was used in the readme @benninkcorien

Have you tried with GPU instead of CPU?

I tried with slow speech ref clips but the results seems not good.

@coolst3r your solution looks like generated by ai, looks like not quite relevant to our question here.

@chenaaron3 Had a try with your method but still error: ``` TypeError: Firefox.__init__() got an unexpected keyword argument 'profile_path' ```

@mb-alex-b-c thank you for your help. Any idea how to speed up the process here? I am using 3080 Ti locally. But observe that GPU workload is just 1%-2% all...

Thanks again for your help. But I do need to run this locally, perhaps with support on GPU.

@Fictiverse Thanks, sorted it out now!

How to train a lora? Much appreciate it!