makemore
makemore copied to clipboard
karpathy
An autoregressive character-level language model for making more things
creating the interface for makemore, so any one identify the path of the data file and play in some customizable settings to adjust them then just clicks on generate without...
First of all, thanks from a 45yo full stack dev that never stop learning and really enjoyed your videos!! I know I'm picky but as a dev I feel the...
Resolves #19 This PR adds a Gradnorm view to TensorBoard by computing the average of the L2 gradnorm across all parame Here's how it appears in TensorBoard: ters. Signed-off-by: Oleg
In the makemore series, the max word length for the basic MLP was 3, which was nice because you could tune hyperparameters much quicker. The current implementation simply calculates the...
We should probably add the gradnorm to the tensorboard plots to see what the convergence is like over time.
Hi Andrej, I wanted to share that, with a few judicious modifications to the Transformer code in makemore.py (in particular, a custom tokenizer as shown in the image + cross...
Hello everyone, i've made a `makemore.py` file (same as https://github.com/karpathy/makemore/blob/master/makemore.py) with only Bigram model class as of now. while running it with `python makemore.py -i names.txt -o names --type bigram`...
