init_model_from argument has no effect on where driver starts

karpathy / neuraltalk

NeuralTalk is a Python+numpy project for learning Multimodal Recurrent Neural Networks that describe images with sentences.

5.41k stars 1.32k forks source link

init_model_from argument has no effect on where driver starts #28

Closed EricZeiberg closed 9 years ago

EricZeiberg commented 9 years ago

It seems like, even when a checkpoint file is passed into --init_model_from argument, it starts from epoch 0.00 and acts like the initial model was never even passed in.

karpathy commented 9 years ago

it's init_from, not resume_from. it loads in the weights but it doesnt actually resume the optimization exactly at the epoch it did before. instead it starts the counter at scratch. i decided that it would be slightly tricky and didn't have time to do it.

EricZeiberg commented 9 years ago

Oh alright