AI/text-generation-webui
1
0
Fork 0
mirror of https://github.com/oobabooga/text-generation-webui.git synced 2024-10-01 01:26:03 -04:00
Code Issues Packages Projects Releases Wiki Activity

Bug fixes

Browse Source
This commit is contained in:
oobabooga 2023-03-08 11:26:29 -03:00
parent 33fb6aed74
commit add9330e5e
1 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
modules/text_generation.py
View File

@ -115,7 +115,8 @@ def generate_reply(question, max_new_tokens, do_sample, temperature, top_p, typi
print(f"\n\n{question}\n--------------------\n")
input_ids = encode(question, max_new_tokens)
original_input_ids = output = input_ids
original_input_ids = input_ids
output = input_ids[0]
cuda = "" if (shared.args.cpu or shared.args.deepspeed or shared.args.flexgen) else ".cuda()"
n = shared.tokenizer.eos_token_id if eos_token is None else int(encode(eos_token)[0][-1])
if stopping_string is not None:
@ -186,7 +187,8 @@ def generate_reply(question, max_new_tokens, do_sample, temperature, top_p, typi
if 'stopping_criteria' not in kwargs:
kwargs['stopping_criteria'] = []
kwargs['stopping_criteria'].append(Stream(callback_func=callback))
shared.model.generate(**kwargs)[0]
clear_torch_cache()
shared.model.generate(**kwargs)
def generate_with_streaming(**kwargs):
return Iteratorize(generate_with_callback, kwargs, callback=None)
@ -208,7 +210,6 @@ def generate_reply(question, max_new_tokens, do_sample, temperature, top_p, typi
else:
for i in range(max_new_tokens//8+1):
clear_torch_cache()
with torch.no_grad():
output = eval(f"shared.model.generate({', '.join(generate_params)})")[0]
if shared.soft_prompt: