diff options
Diffstat (limited to 'README.md')
-rw-r--r-- | README.md | 6 |
1 files changed, 4 insertions, 2 deletions
@@ -75,5 +75,7 @@ Ideas of mine that did not work unfortunately: - [ ] fix linting - [ ] Modularize the decoder - [ ] Add kv cache -- [ ] Fix stems -- [ ] residual attn +- [x] Fix stems +- [x] residual attn +- [x] single kv head +- [x] fix rotary embedding |