原书出处
Build a Reasoning Model (From Scratch)
we can read the output while it's being generated. At the same time, we collect the generated tokens into a list so that we can later decode them into the complete final answer string. This pattern of both streaming and collecting tokens is handy because it lets us monitor the generation live while still storing the full answer text (all_tokens) that we can process later.
原书定位详情
来源:a19c7933-4eda-4176-aebd-30fe7e1ad292
正文位置:extracted/Build a Reasoning Model (From Scratch)/_epub-src/OEBPS/Text/chapter-3.html
EPUB SHA-256:52d4d07986020be70f5aac21762dfcf60fcba08db0ba29d4f63957caf12aa377