FineTuning With multiple targets

[francoisabcd]

Hi,

I was trying to finetune with multiple targets as showing below:
train_inputs.append(
{
"target": {col: df_loop[col].values for col in list_cols_target},
"past_covariates": {
col: df_loop[col].values for col in list_past_covariates
},
}
)
# Fine-tune the model
finetuned_pipeline_2 = pipeline.fit(
inputs=train_inputs,
prediction_length=prediction_length,
num_steps=num_steps,
learning_rate=learning_rate,
batch_size=32,
logging_steps=10,
)

  However, it is not working.
  Is it possible to fine-tune with multiple targets.
  Thank you,

[abdulfatir]

@francoisabcd Could you please provide a working example to reproduce the issue? What do you mean by "not working"? Also, you can use backticks to format your code correctly.

[francoisabcd]

It seems that the issue is here "target": {col: df_loop[col].values for col in list_cols_target}. "target" should be 2-D array.