huggingface/candle
1
0
Fork 0
mirror of https://github.com/huggingface/candle.git synced 2025-06-19 11:56:45 +00:00
Code Issues Packages Projects Releases Wiki Activity

Remove the unused pragma in mpt. (#1122)

Browse Source
This commit is contained in:
Laurent Mazare
2023-10-18 15:47:50 +01:00
committed by GitHub
parent 63c204c79e
commit cb034506cd
1 changed files with 4 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
candle-transformers/src/models/mpt.rs
View File

@ -1,9 +1,8 @@
#![allow(unused)] use crate::models::with_tracing::{linear_no_bias, Embedding, Linear};
use crate::models::with_tracing::{linear_no_bias, Embedding as E, Linear};
/// MPT model used by replit-code-v1_5-3b /// MPT model used by replit-code-v1_5-3b
/// https://huggingface.co/replit/replit-code-v1_5-3b/blob/main/modeling_mpt.py /// https://huggingface.co/replit/replit-code-v1_5-3b/blob/main/modeling_mpt.py
use candle::{DType, Device, IndexOp, Module, Result, Tensor, D}; use candle::{DType, Device, IndexOp, Module, Result, Tensor, D};
use candle_nn::{layer_norm, Activation, LayerNorm, VarBuilder}; use candle_nn::{layer_norm, LayerNorm, VarBuilder};
// https://huggingface.co/replit/replit-code-v1_5-3b/blob/main/configuration_mpt.py // https://huggingface.co/replit/replit-code-v1_5-3b/blob/main/configuration_mpt.py
#[derive(Debug, Clone, PartialEq)] #[derive(Debug, Clone, PartialEq)]
@ -243,14 +242,14 @@ fn build_alibi_bias(cfg: &Config) -> Result<Tensor> {
#[derive(Debug)] #[derive(Debug)]
pub struct Model { pub struct Model {
wte: candle_nn::Embedding, wte: Embedding,
blocks: Vec<MPTBlock>, blocks: Vec<MPTBlock>,
norm_f: LayerNorm, norm_f: LayerNorm,
} }
impl Model { impl Model {
pub fn new(cfg: &Config, vb: VarBuilder) -> Result<Self> { pub fn new(cfg: &Config, vb: VarBuilder) -> Result<Self> {
let wte = candle_nn::embedding(cfg.vocab_size, cfg.d_model, vb.pp("wte"))?; let wte = Embedding::new(cfg.vocab_size, cfg.d_model, vb.pp("wte"))?;
let vb_b = vb.pp("blocks"); let vb_b = vb.pp("blocks");
let mut blocks = Vec::with_capacity(cfg.n_layers); let mut blocks = Vec::with_capacity(cfg.n_layers);
for i in 0..cfg.n_layers { for i in 0..cfg.n_layers {