Pinned Loading

1. minibatch-llm minibatch-llm Public

   From-scratch LLM inference server: static then continuous (iteration-level) batching, with an honest throughput vs latency benchmark.

   Python

2. minimind minimind Public

   Forked from jingyaogong/minimind

   🧠 Train a 64M-parameter LLM from scratch in just 2h!

   Python

3. pytorch-cnn-practice pytorch-cnn-practice Public

   Python

4. SWE-Review-Bench SWE-Review-Bench Public

   Cold code-review bug-finding benchmark on SWE-bench Lite. Pilot run (n=20) with three reviewers, prompt-variant probe, Wilson 95% CIs, and a pytest leakage suite.

   Python

5. text2sql-rlvr text2sql-rlvr Public

   Python

6. verifier-guided-planning-agent verifier-guided-planning-agent Public

   Python