Abi See

> Or you just let the program to determine where to stop the whole summary, which means it actually leans to place comma and period properly. Yes, that's how it's...

Hi everyone, I'm afraid this will be a rather unsatisfactory answer, but the reason that [we set default device to CPU](https://github.com/abisee/pointer-generator/blob/master/run_summarization.py#L125) is that our code is derived from the [TextSum...

Hello again, Though I'm sure I tried changing `default_device` several months ago and found no difference, I tried making the change again this week and it did make things faster....

@LilyZL, currently there is no maximum training step setting but you could easily insert such a constraint by grabbing the variable `train_step` [here](https://github.com/abisee/pointer-generator/blob/master/run_summarization.py#L172) and altering the `while True` loop [here](https://github.com/abisee/pointer-generator/blob/master/run_summarization.py#L155)....

There's a large new grounded dialogue dataset that came out last month: **Conversing by Reading: Contentful Neural Conversation with On-demand Machine Reading** https://arxiv.org/pdf/1906.02738.pdf Might be useful?

I'm still having this problem. `node --version` is `v12.13.1`. I tried doing @tszpinda's suggestion but I still get the same error. Does anyone have any ideas?