mmarch/text-generation-webui
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Small bug fix

Browse source
This commit is contained in:
oobabooga 2023-01-09 12:56:54 -03:00
parent a751d7e693
commit 980f8112a7
1 changed files with 2 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

3
server.py
View file

@ -103,7 +103,7 @@ def generate_reply(question, temperature, max_length, inference_settings, select
if eos_token is None: if eos_token is None:
output = eval(f"model.generate(input_ids, {preset}){cuda}") output = eval(f"model.generate(input_ids, {preset}){cuda}")
else: else:
n = tokenizer.encode(eos_token, return_tensors='pt')[0][1] n = tokenizer.encode(eos_token, return_tensors='pt')[0][-1]
output = eval(f"model.generate(input_ids, eos_token_id={n}, {preset}){cuda}") output = eval(f"model.generate(input_ids, eos_token_id={n}, {preset}){cuda}")
reply = tokenizer.decode(output[0], skip_special_tokens=True) reply = tokenizer.decode(output[0], skip_special_tokens=True)
@ -131,6 +131,7 @@ else:
print(f"{i+1}. {model}") print(f"{i+1}. {model}")
print(f"\nWhich one do you want to load? 1-{len(available_models)}\n") print(f"\nWhich one do you want to load? 1-{len(available_models)}\n")
i = int(input())-1 i = int(input())-1
print()
model_name = available_models[i] model_name = available_models[i]
model, tokenizer = load_model(model_name) model, tokenizer = load_model(model_name)