feat: Add Tiny Char-Level GPT That Learns Toy Sequences - #11
Open
ThomasHartDev wants to merge 1 commit into
Open
feat: Add Tiny Char-Level GPT That Learns Toy Sequences#11ThomasHartDev wants to merge 1 commit into
ThomasHartDev wants to merge 1 commit into
Conversation
One-block decoder-only LM over characters with causal self-attention, hand-derived backprop through attention, and Adam training on a toy next-token task.
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds a one-block decoder-only language model over characters: token + position embeddings, causal self-attention, a residual ReLU FFN, and a linear vocab head, with hand-written backprop through the attention path. Trains with the existing Adam optimizer on next-token cross-entropy so a repeating ABAB corpus is enough to drop loss and prefer the right next character. Generate samples autoregressively under a temperature. pytest covers shapes, causal independence of past logits, finite-difference grads, and the learning check.