Doubts while implementing Extract aspect terms and classify sentiments

[nsvradhamadhavi]

I followed the following Steps:

1. Take a csv file with the reviews without header.
2. Use make_ABSA_dataset() and Auto-annoate your datasets via PyABSA - IT is giving the outputs/resuts :
   The results of aspect term extraction have been saved in /home/radhamadhavi/Aspect Term Extraction and Polarity Classification.FAST_LCF_ATEPC.result.json
   and APC and ATEPC Datasets built for the csv file loaded in step 1
3. and also asking for : need to add ID for your dataset,
   and move the generated datasets to integrated_dataset/apc_datasets and integrated_dataset/atepc_datasets, respectively
4. I moved these two datasets to the folders apc_datasets and atepc_datasets in integrated_datasets folder,
   after creating a folder with some id number for example 3011.mydataset in each folder apc_datasets and atepc_datasets
   integrated_datasets--> apc_datasets-->3011.mydataset-->mydata.train.dat.apc &
   integrated_datasets--> atepc_datasets-->3011.mydataset-->mydata.train.dat.atepc
5. Now my question is: how to get the .inference file for these apc/atepc datasets?
6. How to predict aspect and sentiment for my data which is converted to apc/atepc format,
   using the pyabsa trained model. It is a bit confusing at the line before batch_predict:
   inference_source = ATEPC.ATEPCDatasetList.Restaurant16 . How to give my data here?

I would like to implement the following:

1. Using pretrained models of pyabsa, I would like to predict the aspects and sentiments for my reviews;
2. Using pyabsa package, I would like to train the model using my reviews and predict aspects and sentiments on the train and valid/test datasets.
   Please help in this implementation.
   Please guide me