我创建了我的数据集并转换为 COCO。RetinaNet 已安装在 TPU(官方)中。我想用这个 RetinaNet 用我自己的数据集训练一个模型,但总是得到同样的错误。
export MODEL_DIR=${STORAGE_BUCKET}/retinanet-train-eval; \
export RESNET_CHECKPOINT=gs://cloud-tpu-artifacts/resnet/resnet-nhwc-2018-10-14/model.ckpt-112602; \
export TRAIN_FILE_PATTERN=${STORAGE_BUCKET}/coco/coco_train*; \
export EVAL_FILE_PATTERN=${STORAGE_BUCKET}/coco/coco_val*; \
export VAL_JSON_FILE=${STORAGE_BUCKET}/coco/instances_val.json; \
export EVAL_SAMPLES=613 \
export NUM_STEPS_PER_EVAL=868
python /usr/share/tpu/models/official/detection/main.py \
--use_tpu=True \
--tpu="${TPU_NAME?}" \
--num_cores=8 \
--model_dir="${MODEL_DIR?}" \
--mode="train_and_eval" \
--params_override="{
type: retinanet,
train: {
checkpoint: {
path: $ {
RESNET_CHECKPOINT ?
},
prefix: resnet50 /
},
train_file_pattern: $ {
TRAIN_FILE_PATTERN ?
}
},
eval: {
val_json_file: $ {
VAL_JSON_FILE ?
},
eval_file_pattern: $ {
EVAL_FILE_PATTERN ?
},
eval_samples: $ {
EVAL_SAMPLES ?
},
num_steps_per_eval: $ {
NUM_STEPS_PER_EVAL ?
}
}
}"
使用这些设置,我收到此错误...