mirror of
https://github.com/MLSysBook/TinyTorch.git
synced 2026-04-28 15:22:39 -05:00
Resolved transformer training issues by adding __call__ methods to: - Embedding, PositionalEncoding, EmbeddingLayer (text.embeddings) - LayerNorm, MLP, TransformerBlock, GPT (models.transformer) - MultiHeadAttention (core.attention) This enables PyTorch-style syntax: model(x) instead of model.forward(x) All transformer diagnostic tests now pass (5/5 ✓)(https://claude.ai/code)