Carlos Mocholí

For anybody desperate, this is a valid workaround: ```diff diff --git a/src/lightning/pytorch/trainer/connectors/logger_connector/result.py b/src/lightning/pytorch/trainer/connectors/logger_connector/result.py index 81630362a..4d0b88208 100644 --- a/src/lightning/pytorch/trainer/connectors/logger_connector/result.py +++ b/src/lightning/pytorch/trainer/connectors/logger_connector/result.py @@ -398,10 +398,12 @@ class _ResultCollection(dict): self.register_key(key, meta, value) #...

I would strongly suggest that you don't pickle the Trainer object. This is a bad idea as you are pickling the precise code and imports, which might break with future...

@ioangatop It looks like your inheritance structure violates the liskov substitution principle. Instead of bending the hook system to workaround its limitations, here's two simpler alternatives - set `limit_val_batches=0` to...

The script is designed to stop when you pass an empty line: https://github.com/Lightning-AI/litgpt/blob/main/litgpt/chat/base.py#L173-L174. I suggest that you debug what `input()` gets when you copy paste a piece of text like...

Is that enough? I thought this would also need a way to let it know that you finished typing to support newlines

Can you share the hparams printed in the command line? In case this is a parsing issue

@awaelchli Should we add a `PretrainingDataset` just like we have for SFT? Then the pretrain file can set this as the expected type

> If we wanted to restrict which scripts can use which data modules, we could have base classes for pretraining and fine-tuning. This is what I meant

This piece of validation was useful to subjectively evaluate the output of a fine-tuned model on the alpaca dataset as training progressed. Reusing it is on purpose so that you...

You could do that, but then it might be different if you shuffle or change your dataset splits