Documentation Pass for Models (#2617)

* links in chinese_clip

* links for clip model

* add mod docs for flux and llava

* module doc for MMDIT and MIMI

* add docs for a few more modesl

* mod docs for bert naser and beit

* add module docs for convmixer colpali codegeex and chatglm

* add another series of moddocs

* add  fastvit-llama2_c

* module docs mamba -> mobileone

* module docs from moondream-phi3

* mod docs for quantized and qwen

* update to yi

* fix long names

* Update llama2_c.rs

* Update llama2_c_weights.rs

* Fix the link for mimi + tweaks

---------

Co-authored-by: Laurent Mazare <laurent.mazare@gmail.com>

1 files changed, 17 insertions, 0 deletions

diff --git a/candle-transformers/src/models/vit.rs b/candle-transformers/src/models/vit.rs
index 3be72bf5..49ab4630 100644
--- a/candle-transformers/src/models/vit.rs
+++ b/candle-transformers/src/models/vit.rs
@@ -1,3 +1,20 @@
+//! Vision Transformer (ViT) implementation.
+//!
+//! Vision Transformer applies transformer architecture to image classification
+//! by splitting images into patches and processing them as a sequence.
+//!
+//! Key characteristics:
+//! - Image patches as sequence tokens
+//! - Self-attention between patches
+//! - Position embeddings
+//! - CLS token for classification
+//! - Layer normalization
+//!
+//! References:
+//! - [ViT Paper](https://arxiv.org/abs/2010.11929)
+//! - [Model Card](https://huggingface.co/google/vit-base-patch16-224)
+//!
+
 use crate::models::with_tracing::{conv2d, linear, linear_no_bias, Conv2d, Linear};
 use candle::{IndexOp, Module, Result, Tensor, D};
 use candle_nn::{layer_norm, LayerNorm, VarBuilder};