llama.cpp llama-cli cannot load : unknown pre-tokenizer type: 'trillion'
#1
by
lefromage
- opened
llama-cli -hf trillionlabs/Trillion-7B-preview-GGUF:Q4_K_M
llama_model_load: error loading model: error loading model vocabulary: unknown pre-tokenizer type: 'trillion'
llama_model_load_from_file_impl: failed to load model
Hi
@lefromage
, thanks for your interest in Trillion-7B-preview!
We're working on supporting our tokenizer with llama-cpp, and it should be ready in a few days.
Please stay tuned!
cool! Thanks
This fix has been merged in llama-cpp. You should also be able to access our models through LM Studio after applying the update.
Thanks!
sungjunhan-trl
changed discussion status to
closed