mirror of https://github.com/google/gemma.cpp.git
Only the weights; binding MatMul output worsens batch=1 prefill. Update gemma_batch_bench to use --decode_qbatch. Fix/remove prefill_activations in gemma-inl.h. Refactor: use BasePageBytes directly when binding Move BindB/C to .cc by de-templatizing Remove MatOwners::AllocateFor because it is weights-specific (binding or not) Disband MatOwners, replace with vector PiperOrigin-RevId: 759610477 |
||
|---|---|---|
| .. | ||
| activations.h | ||
| backward-inl.h | ||
| backward.cc | ||
| backward.h | ||
| backward_scalar.h | ||
| backward_test.cc | ||
| common_scalar.h | ||
| forward-inl.h | ||
| forward.cc | ||
| forward.h | ||
| forward_scalar.h | ||
| optimize_test.cc | ||
| optimizer.cc | ||
| optimizer.h | ||
| prompt.h | ||
| sampler.h | ||
| test_util.h | ||