ValueError: Dimension size must be evenly divisible by 3 but is 1
Number of ways to split should evenly divide the split dimension for 'imagetransformer/parallel_0_4/imagetransformer/imagetransformer/body/split' (op: 'Split') with input shapes: [], [?,1,1,1] and with computed input tensors: input[0] = <3>.
Can anyone tell me why the problem happened? Thanks so much.
Sorry everyone. I have known and be able to train the model in MODEL=imagetransformer2d, but still cannot train it in the MODEL=imagetransformer, can anyone help me to solve the problem?
Description
Why cannot I train the model by the imagetransformer?
The command code:
pip install tensor2tensor
PROBLEM=image_cifar10_plain_gen MODEL=imagetransformer HPARAMS=imagetransformer_cifar10_base
DATA_DIR=$HOME/t2t_data TMP_DIR=/tmp/t2t_datagen TRAIN_DIR=$HOME/t2t_train/$PROBLEM/$MODEL-$HPARAMS
mkdir -p $DATA_DIR $TMP_DIR $TRAIN_DIR
t2t-datagen \ --data_dir=$DATA_DIR \ --tmp_dir=$TMP_DIR \ --problem=$PROBLEM
t2t-trainer \ --data_dir=$DATA_DIR \ --problem=$PROBLEM \ --model=$MODEL \ --hparams_set=$HPARAMS \ --output_dir=$TRAIN_DIR
The error:
ValueError: Dimension size must be evenly divisible by 3 but is 1 Number of ways to split should evenly divide the split dimension for 'imagetransformer/parallel_0_4/imagetransformer/imagetransformer/body/split' (op: 'Split') with input shapes: [], [?,1,1,1] and with computed input tensors: input[0] = <3>.
Can anyone tell me why the problem happened? Thanks so much.