optimizer.zero_grad()
def criterion(output, target, steps, batch_size):
loss = F.cross_entropy(output, target)
loss.backward()
return loss
Mariano Kamp marianokamp
dbpprt
/ oom.md
Last active
September 6, 2021 13:27
A simple helper function to handle OOM errors while training with PyTorch. On my Windows system I sometimes get strange OutOfMemory errors in the middle of a training job. This wrapper tries to recover by freeing up as much memory as possible and splits the batches into half.
karpathy
/ min-char-rnn.py
Last active
February 26, 2025 02:03
Minimal character-level language model with a Vanilla Recurrent Neural Network, in Python/numpy
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| """ | |
| Minimal character-level Vanilla RNN model. Written by Andrej Karpathy (@karpathy) | |
| BSD License | |
| """ | |
| import numpy as np | |
| # data I/O | |
| data = open('input.txt', 'r').read() # should be simple plain text file | |
| chars = list(set(data)) | |
| data_size, vocab_size = len(data), len(chars) |
git clone [email protected]:YOUR-USERNAME/YOUR-FORKED-REPO.git
cd into/cloned/fork-repo
git remote add upstream git://github.com/ORIGINAL-DEV-USERNAME/REPO-YOU-FORKED-FROM.git
git fetch upstream