resume a training #250

New Issue

icklerly · 2023-05-25T11:13:01Z

icklerly commented

2023-05-25 11:13:01 +00:00

Hey,

i have a question:
how do I resume a training? Is this possible?

When I click on start training it just generates a new directory of finetung. It starts from epoch 0 again.

Hey, i have a question: how do I resume a training? Is this possible? When I click on start training it just generates a new directory of finetung. It starts from epoch 0 again.

mrq commented

2023-05-25 14:11:44 +00:00

Under the training configuration generator sub-tab:

Resume State Path: the last training state saved to resume from. The general path structure is what the placeholder value is. This will resume from whatever iterations it was last at, and iterate from there until the target step count (for example, resuming from iteration 2500, while requesting 5000 iterations, will iterate 2500 more times).

Under the training configuration generator sub-tab: > Resume State Path: the last training state saved to resume from. The general path structure is what the placeholder value is. This will resume from whatever iterations it was last at, and iterate from there until the target step count (for example, resuming from iteration 2500, while requesting 5000 iterations, will iterate 2500 more times).

👍 1

DoctorPopi commented

2023-09-17 12:09:05 +00:00

Hi! Is it possible to resume a training and add new clips? Or should that rather be another finetuning altogether?

I suppose this line in the wiki answers the question but I'm not sure I understand:

!NOTE!: ensure your dataset size and batch sizes match, as things will get botched if you don't

Hi! Is it possible to resume a training and add new clips? Or should that rather be another finetuning altogether? I suppose this line in the wiki answers the question but I'm not sure I understand: `!NOTE!: ensure your dataset size and batch sizes match, as things will get botched if you don't`

mrq commented

2023-09-17 12:51:56 +00:00

Hi! Is it possible to resume a training and add new clips? Or should that rather be another finetuning altogether?

#353

> Hi! Is it possible to resume a training and add new clips? Or should that rather be another finetuning altogether? https://git.ecker.tech/mrq/ai-voice-cloning/issues/353

DoctorPopi commented

2023-09-17 15:05:01 +00:00

Perfect! Thank you!

Sign in to join this conversation.