In the report I read that 4 V100 GPUs were used for finetuning the models, however, I have not found anything regarding how much time it took. I want the fine-tuned (review comment generation) checkpoint, but as I saw they are not being released. Since the only possibility I have is to train it on CPU, it would be helpful to know how much time it took for you, so that I can estimate the time it would take me on CPU, and reduce the size of the datasets according to that. This question is not only for the creators, so if anyone did the fine-tuning, please let me know what hardware you had and how much time it took. Alternatively, if someone has the fine-tuned checkpoint, I would be really happy if they could send it to me.
In the report I read that 4 V100 GPUs were used for finetuning the models, however, I have not found anything regarding how much time it took. I want the fine-tuned (review comment generation) checkpoint, but as I saw they are not being released. Since the only possibility I have is to train it on CPU, it would be helpful to know how much time it took for you, so that I can estimate the time it would take me on CPU, and reduce the size of the datasets according to that. This question is not only for the creators, so if anyone did the fine-tuning, please let me know what hardware you had and how much time it took. Alternatively, if someone has the fine-tuned checkpoint, I would be really happy if they could send it to me.