add graph-based config and model-- ultragcn #251
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
|
|
CI PY3 Test Passed |
|
CI Test Failed |
tiankongdeguiji
requested changes
Jul 21, 2022
| self._nbr_weights = features.get('features')[5] | ||
| self._neg_ids = features.get('features')[6] | ||
| else: | ||
| self._user_ids = features.get('id') |
| def get_outputs(self): | ||
| # emb_1 = tf.reduce_join(tf.as_string(self._prediction_dict['user_embedding']), axis=-1, separator=',') | ||
| # emb_2 = tf.reduce_join(tf.as_string(self._prediction_dict['item_embedding'] ), axis=-1, separator=',') | ||
| return ['user_embedding','item_embedding'] |
Collaborator
There was a problem hiding this comment.
建议跟向量召回保持一致，user_emb， item_emb
| def build_metric_graph(self, eval_config): | ||
| metric_dict = {} | ||
| for metric in eval_config.metrics_set: | ||
| if metric.WhichOneof('metric') == 'recall_at_topk': |
Collaborator
There was a problem hiding this comment.
metric会生效么？logits来自哪里？
| } | ||
|
|
||
| message GraphLearnInput { | ||
| optional string user_node_input = 1; |
Collaborator
There was a problem hiding this comment.
这些是可枚举的么？node_name, node_input，这种kv的形式是不是通用一些？
| import json | ||
| import logging | ||
|
|
||
| from easy_rec.python.utils import pai_util |
Collaborator
There was a problem hiding this comment.
跟core/sampler.py中的graph init复用
| output_types = [tf.int64, tf.float32, tf.int64, tf.float32, | ||
| tf.int64, tf.float32, tf.int64] | ||
| # user ids, user degrees, item ids, item degrees, nbr item ids, nbr item weight, neg item ids | ||
| output_shapes = [tf.TensorShape([None]), |
Collaborator
There was a problem hiding this comment.
output不建议是list，可读性较差，建议是dict
| break | ||
| epoch_id += 1 | ||
|
|
||
| self._nbr_num = self._data_config.ultra_gcn_sampler.nbr_num |
Collaborator
There was a problem hiding this comment.
不应该修改GraphInput类的成员变量，如果需要建成员变量的话，建议单独建个ultra gcn sampler的类，被GraphInput实例化
| if self._sampler is not None and self._mode != tf.estimator.ModeKeys.PREDICT: | ||
| if self._mode != tf.estimator.ModeKeys.TRAIN: | ||
| self._sampler.set_eval_num_sample() | ||
| sampler_type = self._data_config.WhichOneof('sampler') |
Collaborator
There was a problem hiding this comment.
负采样和feature config还需要么？
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.