Spaces:
Running
on
Zero
Running
on
Zero
Selecting Apollo InternLM doesn't work
#1
by
SaisExperiments
- opened
I guess it's cause the model doesn't exist on HF, but it always ends in "GPU task aborted". I've been waiting for an InternLM finetune to mess with 😶🌫️
Hmm your right, it's an issue with the tokenizer. It's a privated model that I intend to release soon.
Had to play around with the tokenizer config. It works now.
Locutusque
changed discussion status to
closed
I'd also like to mention, you should probably use the llama based model instead as InternLM seems to be broken.