Project structure
Explore the files, run the current app, and see what is verified for a full build.
browser-chat/
0 of 14 lesson files readyThe React app imports real TypeScript modules. Your standalone Python files implement the same host-owned contracts; models/character-rnn.py can also produce the checkpoint the app runs.
- Imported app graph
main.tsx → BrowserChat.tsx → browser adapters- Python contracts
standalone .py files → host-owned checks- Trained model
character-rnn.py → checkpoint → Student RNN- Runtime settings
*.config.js → host-parsed configuration
runtime/
- latent-tensor.jsProvided
- model.config.jsReady to edit
- transport.config.jsReady to edit
- interface.config.jsReady to edit
- host-bridge.tsProvided
- model-softmax.jsProvided
- streaming-transport.jsProvided
- generation-reliability.jsProvided
- chat-reducer.jsProvided
- chat-actions.jsProvided
- chat-quality.jsProvided
- streaming-react.jsProvided
models/
product/
History and privacy
Project history
The full file tree is scaffolded from the start. Move through the course to see its placeholders replaced by verified lesson work.
browser-chat/models/character-rnn.pyScaffolded placeholdermodels/neural-language-model.pyScaffolded placeholdermodels/bpe-tokenizer.pyScaffolded placeholdermodels/additive-attention.pyScaffolded placeholdermodels/causal-transformer.pyScaffolded placeholdermodels/few-shot-evaluation.pyScaffolded placeholdersystems/inference-runtime.pyScaffolded placeholdersystems/continuous-batching.pyScaffolded placeholderbackend/streaming-transport.pyScaffolded placeholderbackend/generation-reliability.pyScaffolded placeholderproduct/chat-reducer.pyScaffolded placeholderproduct/streaming-react.pyScaffolded placeholderproduct/chat-actions.pyScaffolded placeholderproduct/chat-quality.pyScaffolded placeholder
This log contains event names and pass/fail results only. It does not include code, prompts, messages, API keys, or written answers.