Load xgen tokenizer

This commit is contained in:
Honkware 2023-06-29 01:32:44 -05:00
parent c6cae106e7
commit 0a6a498383

View File

@ -94,6 +94,8 @@ def load_tokenizer(model_name, model):
if any(s in model_name.lower() for s in ['gpt-4chan', 'gpt4chan']) and Path(f"{shared.args.model_dir}/gpt-j-6B/").exists():
tokenizer = AutoTokenizer.from_pretrained(Path(f"{shared.args.model_dir}/gpt-j-6B/"))
elif model.__class__.__name__ in ['LlamaForCausalLM', 'LlamaGPTQForCausalLM', 'ExllamaHF']:
if any(s in model_name.lower() for s in ['xgen']) and Path(f"{shared.args.model_dir}/{model_name}/").exists():
tokenizer = AutoTokenizer.from_pretrained(Path(f"{shared.args.model_dir}/{model_name}/"), trust_remote_code=shared.args.trust_remote_code)
# Try to load an universal LLaMA tokenizer
if not any(s in shared.model_name.lower() for s in ['llava', 'oasst']):
for p in [Path(f"{shared.args.model_dir}/llama-tokenizer/"), Path(f"{shared.args.model_dir}/oobabooga_llama-tokenizer/")]: