aim-se/dif-sr

dif-sr's Issues

code?

Could you upload this code?

code

Thank the author for your contribution. After reading your paper, I began to learn your network structure. How to download the implementation code? Thank you for sharing

Typo in title of notebook

DIF-SR/benchmarks/DIF_yelp.md

Line 1 in b600052

## DIF_Amazon_Beauty

Here it should be Yelp I guess.

IndexError: list index out of range

selected_features长度大于1会出现下面错误：

[copy.deepcopy(FeatureSeqEmbLayer(dataset,self.attribute_hidden_size[],[self.selected_features[]],self.pooling_mode,self.device)) for _
IndexError: list index out of range

where is the AAP implement

The AAP operation mentioned in the article, I don't seem to see it implemented in code. Can you give me the specific part where it is implemented, and can you give me the specific code?

Failure to use multi-GPU setting

Hello, thank you for the code you posted.

There is a multi-GPU setting in the code, so I set the config like below

...
gpu_id: '0,1,2,3'
multi_gpus: True
...

In order to use multi-gpu, I used torchrun to run. I put the command as shown below.

torchrun run_recbole.py --model=SASRecD --dataset='Amazon_Beauty' --config_files='configs/Amazon_Beauty.yaml' --nproc=4

However, this kind of error occurs, which doesn't happen in a single GPU setting.

Can you tell me how can I use multi-gpu setting properly?
Does Multi-GPU setting not finished yet?

providing interaction features while training

Thank you for the great code that you provide.

I wanted to give a float-type feature like a rating in the training step.
I checked your code and it seems it only considers token-type features.
Did I understand correctly?

Thank you

About AAP

Hi, Thanks for your great work.

I would like to ask about the specific label used by AAP in DIF-SR ?
For example, in the Beauty dataset, I print out the categories label of each sample through the code, which is a sample corresponding to a 14-dimensional categories.
Does this ‘categories’ refer to the multi-level categories of the target item?
Whether the category used in the sequence level is a single value or the same 14 dimension value ？
And why ？

question about dataset

hi,
Thank you for providing such good work！
The download url address of the current datasets has a lot of new data. The Amazon and Yelp datasets include different versions in the link. which version of the datasets are you using? Can you provide a more detailed url?
Thank you!

error: when use amazon beauty datasets

when I use amazon beauty datasets
error:

Traceback (most recent call last):
File "run_recbole.py", line 36, in
run_recbole(model=args.model, dataset=args.dataset, config_file_list=config_file_list,config_dict=parameter_dict)
File "/mnt/workspace/myf/DIF_SR/recbole/quick_start/quick_start.py", line 60, in run_recbole
model = get_model(config['model'])(config, train_data.dataset).to(config['device'])
File "/mnt/workspace/myf/DIF_SR/recbole/model/sequential_recommender/sasrecd.py", line 61, in init
in range(len(self.selected_features))])
File "/mnt/workspace/myf/DIF_SR/recbole/model/sequential_recommender/sasrecd.py", line 60, in
[copy.deepcopy(FeatureSeqEmbLayer(dataset,self.attribute_hidden_size[],[self.selected_features[]],self.pooling_mode,self.device)) for _
File "/home/pai/lib/python3.6/copy.py", line 180, in deepcopy
y = _reconstruct(x, memo, *rv)
File "/home/pai/lib/python3.6/copy.py", line 280, in _reconstruct
state = deepcopy(state, memo)
File "/home/pai/lib/python3.6/copy.py", line 150, in deepcopy
y = copier(x, memo)
File "/home/pai/lib/python3.6/copy.py", line 240, in _deepcopy_dict
y[deepcopy(key, memo)] = deepcopy(value, memo)
File "/home/pai/lib/python3.6/copy.py", line 180, in deepcopy
y = _reconstruct(x, memo, *rv)
File "/home/pai/lib/python3.6/copy.py", line 280, in _reconstruct
state = deepcopy(state, memo)
File "/home/pai/lib/python3.6/copy.py", line 150, in deepcopy
y = copier(x, memo)
File "/home/pai/lib/python3.6/copy.py", line 240, in _deepcopy_dict
y[deepcopy(key, memo)] = deepcopy(value, memo)
File "/home/pai/lib/python3.6/copy.py", line 180, in deepcopy
y = _reconstruct(x, memo, *rv)
File "/home/pai/lib/python3.6/copy.py", line 280, in _reconstruct
state = deepcopy(state, memo)
File "/home/pai/lib/python3.6/copy.py", line 150, in deepcopy
y = copier(x, memo)
File "/home/pai/lib/python3.6/copy.py", line 240, in _deepcopy_dict
y[deepcopy(key, memo)] = deepcopy(value, memo)
File "/home/pai/lib/python3.6/copy.py", line 150, in deepcopy
y = copier(x, memo)
File "/home/pai/lib/python3.6/copy.py", line 215, in _deepcopy_list
append(deepcopy(a, memo))
File "/home/pai/lib/python3.6/copy.py", line 180, in deepcopy
y = _reconstruct(x, memo, *rv)
File "/home/pai/lib/python3.6/copy.py", line 280, in _reconstruct
state = deepcopy(state, memo)
File "/home/pai/lib/python3.6/copy.py", line 150, in deepcopy
y = copier(x, memo)
File "/home/pai/lib/python3.6/copy.py", line 240, in _deepcopy_dict
y[deepcopy(key, memo)] = deepcopy(value, memo)
File "/home/pai/lib/python3.6/copy.py", line 169, in deepcopy
rv = reductor(4)
TypeError: can't pickle _thread.RLock objects