Mirrors/text-generation-webui
1
0
Fork 0
mirror of https://github.com/oobabooga/text-generation-webui.git synced 2024-11-22 08:07:56 +01:00
Code Issues Packages Projects Releases Wiki Activity

Small changes

Browse Source
This commit is contained in:
oobabooga 2023-03-16 13:34:23 -03:00
parent 83cb20aad8
commit e085cb4333
1 changed files with 8 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
modules/models.py
View File

@ -7,8 +7,9 @@ from pathlib import Path
import numpy as np
import torch
import transformers
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig, AutoConfig
from accelerate import infer_auto_device_map, init_empty_weights, load_checkpoint_and_dispatch
from accelerate import infer_auto_device_map, init_empty_weights
from transformers import (AutoConfig, AutoModelForCausalLM, AutoTokenizer,
BitsAndBytesConfig)
import modules.shared as shared
@ -113,8 +114,8 @@ def load_model(model_name):
if shared.args.gpu_memory:
memory_map = shared.args.gpu_memory
max_memory = { 0: f'{memory_map[0]}GiB' }
for i in range(1, len(memory_map)):
max_memory = {}
for i in range(len(memory_map)):
max_memory[i] = f'{memory_map[i]}GiB'
max_memory['cpu'] = f'{shared.args.cpu_memory or 99}GiB'
params['max_memory'] = max_memory
@ -126,10 +127,7 @@ def load_model(model_name):
suggestion = int(round(suggestion/1000))
print(f"\033[1;32;1mAuto-assiging --gpu-memory {suggestion} for your GPU to try to prevent out-of-memory errors.\nYou can manually set other values.\033[0;37;0m")
max_memory = {
0: f'{suggestion}GiB',
'cpu': f'{shared.args.cpu_memory or 99}GiB'
}
max_memory = {0: f'{suggestion}GiB', 'cpu': f'{shared.args.cpu_memory or 99}GiB'}
params['max_memory'] = max_memory
if shared.args.disk: