https://huggingface.co/ayan4m1/Clara-v2-8B

#2572
by ayan4m1 - opened

Thanks for your efforts for the community!

https://huggingface.co/ayan4m1/Clara-v2-8B

It's queued!

You can check for progress at http://hf.tst.eu/status.html or regularly check the model
summary page at https://hf.tst.eu/model#Clara-v2-8B-GGUF for quants to appear.

Thank you!

ayan4m1 changed discussion status to closed
ayan4m1 changed discussion status to open

Is the queue running longer than 2 days right now?

Something is wrong with your tokenizer, model tokenizer doesnt seem to be compatible with llamacpp ? Did you use existing tokenizer from base model or made your own tokenizer?

Existing tokenizer from base model.

I was getting errors trying to use llama.cpp quantize scripts on it myself.

https://github.com/unslothai/unsloth/issues/6309#event-26911159756 might be relevant

I was getting errors trying to use llama.cpp quantize scripts on it myself.

we are using the main llamacpp, no forks, no unsloths. I assume something is wrong with the tokenizer, not sure what and cannot try to fix manually due to not having time or abilities to do so. Perhaps try copying the existing tokenizer from original model into your repository, as we did in fact manage to quantize it

Should I copy over tokenizer_config.json as well? Thanks

most probably yes

Done, thank you for all your time and effort!

alright, let me try to requeue

it's quanting =)

Praise be unto the quant.

ayan4m1 changed discussion status to closed

Sign up or log in to comment