ranking
ranking copied to clipboard
tensorflow
Learning to Rank in TensorFlow
I train_and_evaled the model like this: tf.estimator.train_and_evaluate(rank_estimator, train_spec, vali_spec) and got a result metric/ordered_pair_accuracy = 1.0 on the validation set. But, the problem seems that when I ran rank_estimator.predict(input_fn=vali_input_fn) alone...
In the latest release of [Tensorflow Ranking v0.2.3](https://github.com/tensorflow/ranking/releases/tag/v0.2.3), we introduced RankingPipeline, with the goal of minimizing the effort to set up the training workflow for a ranking model. The RankingPipeline...
Example_features are made of tensors of shape (batch_size, list_size, feature_space). If examples are passed in a predictable order (ex. ranked from most relevant to least relevant) the model might not...
Hello, I was trying to use `tfr.keras.metrics.MRRMetric()` at compile time. But it crashes with a shape-related bug. Below is the code I used: ``` mrr_metric = tfr.keras.metrics.MRRMetric(topn=10) def create_classifier_bert_model(): inputs...
Hi TF Ranking team, What is the correct usage of tfr.keras.losses.ClickEMLoss? Can it simply be passed to `model.compile` like others such as NDCG? Thanks!
Hello all, I was following [this example](https://github.com/tensorflow/ranking/blob/master/tensorflow_ranking/examples/keras/keras_dnn_tfrecord.py) and adopting this approach to the implicit recommender system. In my original dataset, there are only instances of positive user interactions with items....
Can I have an example to run distributed ranking on a ray cluster? The example provided is not running on ray cluster can we have an example to run on...
Issue type Bug I'm noticing inconsistent behavior in `tfr.keras.losses.PairwiseHingeLoss` when using labels with negative values (which are masked) or when using precise masks with default reduction. Expectation is to have...
Current behavior: kwargs are passed to tf.keras.Sequential.add, so they are not passed on to tf.keras.layers.Dense as intended. For example, when passing `use_bias=False` to create_tower with the kwarg name `kernel_regularizer`, it...
The example on https://www.tensorflow.org/ranking doesn't run succesfully out of the box. I tried opening it on colab and running, and I get this: ``` Downloading and preparing dataset 1.15 GiB...
