diff options
Diffstat (limited to 'candle-transformers/src/models/mixformer.rs')
| -rw-r--r-- | candle-transformers/src/models/mixformer.rs | 7 |
1 files changed, 7 insertions, 0 deletions
diff --git a/candle-transformers/src/models/mixformer.rs b/candle-transformers/src/models/mixformer.rs index 700829e3..2c2909c3 100644 --- a/candle-transformers/src/models/mixformer.rs +++ b/candle-transformers/src/models/mixformer.rs @@ -1,3 +1,10 @@ +//! MixFormer (Microsoft's Phi Architecture) +//! +//! See "Textbooks Are All You Need II: phi-1.5 technical report", Lin et al. 2023 +//! - [Arxiv](https://arxiv.org/abs/2309.05463) +//! - [Github](https://huggingface.co/microsoft/phi-1_5) +//! + use crate::models::with_tracing::{linear, Embedding as E, Linear}; /// MixFormer model. /// https://huggingface.co/microsoft/phi-1_5 |