我尝试使用tf.estimator.train_and_evaluate()方法从checkpoint中使用已经.config文件重新训练TF对象检测API模型以获取训练管道,例如models / research / object_detection / model_main.py。它每N步或每N秒节省一次检查点。
但我想只保存一个像Keras一样的最佳模型。有没有办法用TF Object Detection API模型做到这一点?也许某些选项/回调用于tf.Estimator.train或某种方式使用Check API与Keras?
你可以尝试使用BestExporter
。据我所知,这是你想要做的唯一选择。
exporter = tf.estimator.BestExporter(
compare_fn=_loss_smaller,
exports_to_keep=5)
eval_spec = tf.estimator.EvalSpec(
input_fn,
steps,
exporters)
https://www.tensorflow.org/api_docs/python/tf/estimator/BestExporter
我一直在使用适合我的https://github.com/bluecamel/best_checkpoint_copier。
例:
best_copier = BestCheckpointCopier(
name='best', # directory within model directory to copy checkpoints to
checkpoints_to_keep=10, # number of checkpoints to keep
score_metric='metrics/total_loss', # metric to use to determine "best"
compare_fn=lambda x,y: x.score < y.score, # comparison function used to determine "best" checkpoint (x is the current checkpoint; y is the previously copied checkpoint with the highest/worst score)
sort_key_fn=lambda x: x.score,
sort_reverse=False) # sort order when discarding excess checkpoints
将它传递给你的eval_spec:
eval_spec = tf.estimator.EvalSpec(
...
exporters=best_copier,
...)