mirror of
https://github.com/oobabooga/text-generation-webui.git
synced 2024-10-01 01:26:03 -04:00
Proper way to free the cuda cache
This commit is contained in:
parent
03d25c1c61
commit
fa58fd5559
@ -1,3 +1,4 @@
|
||||
import gc
|
||||
import re
|
||||
import time
|
||||
|
||||
@ -73,6 +74,8 @@ def formatted_outputs(reply, model_name):
|
||||
return reply
|
||||
|
||||
def generate_reply(question, max_new_tokens, do_sample, temperature, top_p, typical_p, repetition_penalty, top_k, min_length, no_repeat_ngram_size, num_beams, penalty_alpha, length_penalty, early_stopping, eos_token=None, stopping_string=None):
|
||||
gc.collect()
|
||||
if not shared.args.cpu:
|
||||
torch.cuda.empty_cache()
|
||||
|
||||
original_question = question
|
||||
|
Loading…
Reference in New Issue
Block a user