I merged huggyllama/llama-30b + kaiokendev/SuperCOT-LoRA
I forgot to increase the shard size thats why there is so many bin files. I am currently quantizing it to 4bit 128g at the time of writing this.
I merged huggyllama/llama-30b + kaiokendev/SuperCOT-LoRA
I forgot to increase the shard size thats why there is so many bin files. I am currently quantizing it to 4bit 128g at the time of writing this.