File size: 341 Bytes
3efa7b5 2da1b07 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 |
---
license: mit
language:
- zh
- en
pipeline_tag: text-generation
tags:
- agent
---
# YModel1.1
## structure
- using SnifferCaptain's LoE (lack of expert) layer as feed forward.
- using SnifferCaptain's PEGA (Position Embedding Gate Attention) as Transformer attention layer
- using additional identity link between ffn's intermediate part. |