Error when finetuning pretrained huggingface conv-ai chatbot model

I’ve been trying to train the model using my own custom dataset, but the kernel terminates the process just before running the first epoch. The same thing happens on both my local machine, and on Google colab.
The code I run is: ! python train.py --dataset_path counsel_chat_250-tokens_full.json --gradient_accumulation_steps=4 --lm_coef=2.0 --max_history=1 --n_epochs=3 --num_candidates=4 --train_batch_size=2

@julien-c any thoughts?

No sorry :confused:

To be honest that code is pretty ancient now.

1 Like