diff options
Diffstat (limited to 'candle-transformers/src/models/quantized_phi3.rs')
| -rw-r--r-- | candle-transformers/src/models/quantized_phi3.rs | 15 |
1 files changed, 15 insertions, 0 deletions
diff --git a/candle-transformers/src/models/quantized_phi3.rs b/candle-transformers/src/models/quantized_phi3.rs index 257ad983..51a75f38 100644 --- a/candle-transformers/src/models/quantized_phi3.rs +++ b/candle-transformers/src/models/quantized_phi3.rs @@ -1,3 +1,18 @@ +//! Phi3 model implementation with quantization support. +//! +//! Phi3 is a language model intended for research purposes. +//! This implementation provides quantization for reduced memory usage. +//! +//! Key characteristics: +//! - Multi-head attention +//! - RMSNorm for layer normalization +//! - Rotary positional embeddings (RoPE) +//! - Support for quantization +//! +//! References: +//! - [Model Card](https://huggingface.co/microsoft/phi-3) +//! + use std::collections::HashMap; use candle::quantized::gguf_file; |