how to get each user's recommendation items list trained with multi-gpu?

[ucasiggcas]

hi,dear
have seen the NCF ,but a little confused，
after training How to get the items list for every user ?

thx

[scotthoule]

Hi ucasiggcas,

Do you mean how to run the trained model for inference?
It would be great if you could list your specific confusions and have a pointer to the code that doesn't satisfy your need.

Thx

[ucasiggcas]

yes
the target for recommendation is to give the users the items they like,
such as
user_1:[item_1,item_12,item_102,item_9]
user_2:[item_2,item_43,item_34]
....
i think it's not difficult for dev,but No API for this .
so could you pls help me ?

[ucasiggcas]

hi,dear
how to download the datasets quickly?
when I use the down, waiting for a long time, get nothing.
$ python movielens.py

[ucasiggcas]

and got the error down, what's up?

<BisectionDataConstructor(Thread-1, initial daemon)>
General:
  Num users: 138493
  Num items: 26744

Training:
  Positive count:          19861770
  Batch size:              99000 
  Batch count per epoch:   1004

Eval:
  Positive count:          138493
  Batch size:              99000 
  Batch count per epoch:   1399
Traceback (most recent call last):
  File "ncf_keras_main.py", line 549, in <module>
    app.run(main)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/absl/app.py", line 303, in run
    _run_main(main, args)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/absl/app.py", line 251, in _run_main
    sys.exit(main(argv))
  File "ncf_keras_main.py", line 544, in main
    logging.info("Result is %s", run_ncf(FLAGS))
  File "ncf_keras_main.py", line 253, in run_ncf
    params, producer, input_meta_data, strategy))
  File "/data/logs/xulm1/NCF/ncf_input_pipeline.py", line 139, in create_ncf_input_data
    (1 + rconst.NUM_EVAL_NEGATIVES)))
ValueError: Evaluation batch size must be divisible by 2 times 1000

my cmd is as follows
$ python ncf_keras_main.py --model_dir=./tmp/models --data_dir=./tmp/movielens-data --dataset=ml-20m --num_gpus=2

[ucasiggcas]

when I set --num_gpus=1
got another error

Train on None steps, validate on 1399 steps
Epoch 1/2
2020-11-24 15:53:21.501988: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library libcublas.so.10.0
Traceback (most recent call last):
  File "ncf_keras_main.py", line 549, in <module>
    app.run(main)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/absl/app.py", line 303, in run
    _run_main(main, args)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/absl/app.py", line 251, in _run_main
    sys.exit(main(argv))
  File "ncf_keras_main.py", line 544, in main
    logging.info("Result is %s", run_ncf(FLAGS))
  File "ncf_keras_main.py", line 313, in run_ncf
    verbose=2)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training.py", line 727, in fit
    use_multiprocessing=use_multiprocessing)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training_distributed.py", line 685, in fit
    steps_name='steps_per_epoch')
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/tensorflow_core/python/keras/engine/training_arrays.py", line 300, in model_iteration
    batch_outs = f(actual_inputs)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/tensorflow_core/python/keras/backend.py", line 3476, in __call__
    run_metadata=self.run_metadata)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/tensorflow_core/python/client/session.py", line 1472, in __call__
    run_metadata_ptr)
tensorflow.python.framework.errors_impl.InvalidArgumentError: 2 root error(s) found.
  (0) Invalid argument: You must feed a value for placeholder tensor 'user_id' with dtype int32 and shape [?,1]
	 [[{{node user_id}}]]
	 [[loss_layer_1/sparse_categorical_crossentropy/weighted_loss/broadcast_weights/assert_broadcastable/is_valid_shape/has_valid_nonscalar_shape/has_invalid_dims/concat/_297]]
  (1) Invalid argument: You must feed a value for placeholder tensor 'user_id' with dtype int32 and shape [?,1]
	 [[{{node user_id}}]]
0 successful operations.
0 derived errors ignored.

[ucasiggcas]

when I use ml-20m dataset, error

Traceback (most recent call last):
  File "ncf_keras_main.py", line 552, in <module>
    app.run(main)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/absl/app.py", line 303, in run
    _run_main(main, args)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/absl/app.py", line 251, in _run_main
    sys.exit(main(argv))
  File "ncf_keras_main.py", line 547, in main
    logging.info("Result is %s", run_ncf(FLAGS))
  File "ncf_keras_main.py", line 235, in run_ncf
    num_users, num_items, _, _, producer = ncf_common.get_inputs(params)
  File "/data/logs/xulm1/NCF/ncf_common.py", line 47, in get_inputs
    deterministic=FLAGS.seed is not None)
  File "/data/logs/xulm1/NCF/data_preprocessing.py", line 218, in instantiate_pipeline
    raw_data, _ = _filter_index_sort(raw_rating_path, cache_path)
  File "/data/logs/xulm1/NCF/data_preprocessing.py", line 161, in _filter_index_sort
    user_map, item_map, df = read_dataframe(raw_rating_path)
  File "/data/logs/xulm1/NCF/data_preprocessing.py", line 62, in read_dataframe
    grouped = df.groupby(movielens.USER_COLUMN)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/pandas/core/generic.py", line 7894, in groupby
    **kwargs
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/pandas/core/groupby/groupby.py", line 2522, in groupby
    return klass(obj, by, **kwds)
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/pandas/core/groupby/groupby.py", line 391, in __init__
    mutated=self.mutated,
  File "/data/logs/xulm1/myconda/lib/python3.7/site-packages/pandas/core/groupby/grouper.py", line 621, in _get_grouper
    raise KeyError(gpr)
KeyError: 'user_id'

[laxmareddyp]

Hi @ucasiggcas,

This question is better asked on StackOverflow and TensorFlow Forum since it is not a bug or feature request. There is also a larger community that reads questions there.

Thanks

[github-actions]

This issue has been marked stale because it has no recent activity since 7 days. It will be closed if no further activity occurs. Thank you.

[github-actions]

This issue was closed due to lack of activity after being marked stale for past 7 days.

[google-ml-butler]

Are you satisfied with the resolution of your issue?
Yes
No