Train batch size, 表示一次处理几张图片,越大训练时间越短;根据内存大小可扩大数字,一般机器建议1 ...
EPOCH = 1 # train the training data 分享1赞 stablediffusion吧 🌀机制彭于晏🌀 求助大佬,使用kohya训练lora模型一直各种报错Folder 100_liangzai: 2200 steps max_train_steps = 2200 stop_text_encoder_training = 0 lr_warmup_steps = 220 accelerate launch --num_cpu_threads_per_process=2 "train_...
@@ -453,45 +453,47 @@ model.fit(x_train, y_train, batch_size=128, epochs=20, validation_data=(x_val, y ``` Epoch 1/20 -125/125 [===] - 8s 59ms/step - loss: 3.0124 - acc: 0.0669 - val_loss: 2.9522 - val_acc: 0.0725 +125/125 [===] - 8s 57ms/step - loss: ...
实话实说,SD webui的版本就不支持mac系列的卡。炼丹以我经验看,必须买N卡,其他卡都不考虑。社区方...
虽然理论上来说mac和a卡都可以用,但是实际操作SD只能用n卡,我家里有M1max的macstudio和6600xt的双...