AutoML Entity Recognition: None BatchPredict could not start due to empty input CSV file.

[Tom rOB]

I want to do a batch prediction and use the code as described in the documentation

´´´

project_id = "photovoltaicclassifier" 

model_id = "TEN8533967251122946048" 

input_uri = "gs://pv_patents/entity_batch_prediction.jsonl" 

output_uri = "gs://pv_patents/"

```

And my JSONL-File looks like the attachment

Since this method worked great for document classification, I didn't expect any difficulties now. However, I now receive the following error message:

```

GoogleAPICallError Traceback (most recent call last) <ipython-input-7-3bc382f3b219> in <module> 21 print("Waiting for operation to complete...") 22 print( ---> 23 f"Batch Prediction results saved to Cloud Storage bucket. {response.result()}" 24 ) ~\anaconda3\lib\site-packages\google\api_core\future\polling.py in result(self, timeout, retry) 135 # pylint: disable=raising-bad-type 136 # Pylint doesn't recognize that this is valid in this case. --> 137 raise self._exception 138 139 return self._result GoogleAPICallError: None BatchPredict could not start due to empty input CSV file.

```

Especially I wonder about the reference to the CSV file, because I have a JSON file as input.

So what am i doing wrong?

[Tom rOB]

Apparently there were difficulties with the formatting of the JSONL file. I can't say exactly why, but since I recreated the file it works.
But beyond that the error message is a bit misleading.

[Jun Lu]

Hi, 

Google Groups are reserved for general product discussion, you shall consider reaching to StackOverflow for technical questions whereas Issue Tracker for unexpected behaviors and feature requests. 

To get better support you should post to the relevant forum, thus please read the Community Support article for better understanding.