Tracing for the phi model (huggingface#936)

* Add some tracing bits to mixformers.

* Add the missing file.

* Add the conv2d layer to with-tracing.

* Improve the tracing usage.

candle-examples/examples/phi/main.rs

@@ -70,7 +70,7 @@ impl TextGeneration {
         }
         let dt = start_gen.elapsed();
         println!(
-            "{sample_len} tokens generated ({:.3} token/s)",
+            "\n{sample_len} tokens generated ({:.2} token/s)",
             sample_len as f64 / dt.as_secs_f64(),
         );
         Ok(())
@@ -84,6 +84,10 @@ struct Args {
     #[arg(long)]
     cpu: bool,
 
+    /// Enable tracing (generates a trace-timestamp.json file).
+    #[arg(long)]
+    tracing: bool,
+
     #[arg(long)]
     prompt: String,
 
@@ -114,8 +118,19 @@ struct Args {
 }
 
 fn main() -> Result<()> {
+    use tracing_chrome::ChromeLayerBuilder;
+    use tracing_subscriber::prelude::*;
+
     let args = Args::parse();
 
+    let _guard = if args.tracing {
+        let (chrome_layer, guard) = ChromeLayerBuilder::new().build();
+        tracing_subscriber::registry().with(chrome_layer).init();
+        Some(guard)
+    } else {
+        None
+    };
+
     let start = std::time::Instant::now();
     let api = Api::new()?;
     let repo = api.repo(Repo::with_revision(

candle-transformers/src/models/mixformer.rs

@@ -1,3 +1,4 @@
+use crate::models::with_tracing::{linear, Embedding as E, Linear};
 /// MixFormer model.
 /// https://huggingface.co/microsoft/phi-1_5
 /// https://arxiv.org/abs/2309.05463
@@ -58,12 +59,12 @@ impl Config {
 
 #[derive(Debug)]
 struct Embedding {
-    wte: candle_nn::Embedding,
+    wte: E,
 }
 
 impl Embedding {
     fn new(cfg: &Config, vb: VarBuilder) -> Result<Self> {
-        let wte = candle_nn::embedding(cfg.vocab_size, cfg.n_embd, vb.pp("wte"))?;
+        let wte = E::new(cfg.vocab_size, cfg.n_embd, vb.pp("wte"))?;
         Ok(Self { wte })
     }
 }
@@ -143,16 +144,16 @@ impl RotaryEmbedding {
 #[derive(Debug)]
 #[allow(clippy::upper_case_acronyms)]
 struct MLP {
-    fc1: candle_nn::Linear,
-    fc2: candle_nn::Linear,
+    fc1: Linear,
+    fc2: Linear,
     act: Activation,
 }
 
 impl MLP {
     fn new(cfg: &Config, vb: VarBuilder) -> Result<Self> {
         let n_inner = cfg.n_inner.unwrap_or(4 * cfg.n_embd);
-        let fc1 = candle_nn::linear(cfg.n_embd, n_inner, vb.pp("fc1"))?;
-        let fc2 = candle_nn::linear(n_inner, cfg.n_embd, vb.pp("fc2"))?;
+        let fc1 = linear(cfg.n_embd, n_inner, vb.pp("fc1"))?;
+        let fc2 = linear(n_inner, cfg.n_embd, vb.pp("fc2"))?;
         Ok(Self {
             fc1,
             fc2,
@@ -170,13 +171,13 @@ impl Module for MLP {
 #[derive(Debug)]
 struct CausalLMHead {
     ln: candle_nn::LayerNorm,
-    linear: candle_nn::Linear,
+    linear: Linear,
 }
 
 impl CausalLMHead {
     fn new(cfg: &Config, vb: VarBuilder) -> Result<Self> {
         let ln = candle_nn::layer_norm(cfg.n_embd, cfg.layer_norm_epsilon, vb.pp("ln"))?;
-        let linear = candle_nn::linear(cfg.n_embd, cfg.vocab_size, vb.pp("linear"))?;
+        let linear = linear(cfg.n_embd, cfg.vocab_size, vb.pp("linear"))?;
         Ok(Self { ln, linear })
     }
 }
@@ -192,20 +193,21 @@ impl Module for CausalLMHead {
 #[derive(Debug)]
 #[allow(clippy::upper_case_acronyms)]
 struct MHA {
-    wqkv: candle_nn::Linear,
-    out_proj: candle_nn::Linear,
+    wqkv: Linear,
+    out_proj: Linear,
     rotary_emb: RotaryEmbedding,
     kv_cache: Option<(Tensor, Tensor)>,
     head_dim: usize,
     softmax_scale: f64,
+    span: tracing::Span,
 }
 
 impl MHA {
     fn new(cfg: &Config, vb: VarBuilder) -> Result<Self> {
         let head_dim = cfg.n_embd / cfg.n_head;
         let op_size = cfg.n_embd;
-        let wqkv = candle_nn::linear(cfg.n_embd, 3 * op_size, vb.pp("Wqkv"))?;
-        let out_proj = candle_nn::linear(op_size, cfg.n_embd, vb.pp("out_proj"))?;
+        let wqkv = linear(cfg.n_embd, 3 * op_size, vb.pp("Wqkv"))?;
+        let out_proj = linear(op_size, cfg.n_embd, vb.pp("out_proj"))?;
         let rotary_emb = RotaryEmbedding::new(cfg.rotary_dim, MAX_SEQ_LEN, vb.device())?;
         let softmax_scale = 1f64 / (head_dim as f64).sqrt();
         Ok(Self {
@@ -215,10 +217,12 @@ impl MHA {
             kv_cache: None,
             rotary_emb,
             softmax_scale,
+            span: tracing::span!(tracing::Level::TRACE, "mha"),
         })
     }
 
     fn forward(&mut self, xs: &Tensor) -> Result<Tensor> {
+        let _enter = self.span.enter();
         let (b_size, seq_len, _n_embd) = xs.dims3()?;
         let qkv = self
             .wqkv
@@ -267,17 +271,24 @@ struct ParallelBlock {
     ln: candle_nn::LayerNorm,
     mixer: MHA,
     mlp: MLP,
+    span: tracing::Span,
 }
 
 impl ParallelBlock {
     fn new(cfg: &Config, vb: VarBuilder) -> Result<Self> {
         let ln = candle_nn::layer_norm(cfg.n_embd, cfg.layer_norm_epsilon, vb.pp("ln"))?;
         let mixer = MHA::new(cfg, vb.pp("mixer"))?;
         let mlp = MLP::new(cfg, vb.pp("mlp"))?;
-        Ok(Self { ln, mixer, mlp })
+        Ok(Self {
+            ln,
+            mixer,
+            mlp,
+            span: tracing::span!(tracing::Level::TRACE, "block"),
+        })
     }
 
     fn forward(&mut self, xs: &Tensor) -> Result<Tensor> {
+        let _enter = self.span.enter();
         let residual = xs;
         let xs = xs.apply(&self.ln)?;
         let attn_outputs = self.mixer.forward(&xs)?;
@@ -291,6 +302,7 @@ pub struct MixFormerSequentialForCausalLM {
     embedding: Embedding,
     blocks: Vec<ParallelBlock>,
     head: CausalLMHead,
+    span: tracing::Span,
 }
 
 impl MixFormerSequentialForCausalLM {
@@ -307,10 +319,12 @@ impl MixFormerSequentialForCausalLM {
             embedding,
             blocks,
             head,
+            span: tracing::span!(tracing::Level::TRACE, "mixformer"),
         })
     }
 
     pub fn forward(&mut self, xs: &Tensor) -> Result<Tensor> {
+        let _enter = self.span.enter();
         let (_b_size, seq_len) = xs.dims2()?;
         let mut xs = xs.apply(&self.embedding)?;
         for block in self.blocks.iter_mut() {

candle-transformers/src/models/mod.rs

@@ -11,4 +11,5 @@ pub mod segment_anything;
 pub mod stable_diffusion;
 pub mod t5;
 pub mod whisper;
+pub mod with_tracing;
 pub mod wuerstchen;

candle-transformers/src/models/stable_diffusion/resnet.rs

@@ -4,7 +4,7 @@
 //!
 //! Denoising Diffusion Implicit Models, K. He and al, 2015.
 //! https://arxiv.org/abs/1512.03385
-use super::utils::{conv2d, Conv2d};
+use crate::models::with_tracing::{conv2d, Conv2d};
 use candle::{Result, Tensor, D};
 use candle_nn as nn;
 use candle_nn::Module;

candle-transformers/src/models/stable_diffusion/unet_2d.rs

@@ -4,7 +4,7 @@
 //! timestep and return a denoised version of the input.
 use super::embeddings::{TimestepEmbedding, Timesteps};
 use super::unet_2d_blocks::*;
-use super::utils::{conv2d, Conv2d};
+use crate::models::with_tracing::{conv2d, Conv2d};
 use candle::{Result, Tensor};
 use candle_nn as nn;
 use candle_nn::Module;

candle-transformers/src/models/stable_diffusion/unet_2d_blocks.rs

@@ -4,7 +4,7 @@ use super::attention::{
     AttentionBlock, AttentionBlockConfig, SpatialTransformer, SpatialTransformerConfig,
 };
 use super::resnet::{ResnetBlock2D, ResnetBlock2DConfig};
-use super::utils::{conv2d, Conv2d};
+use crate::models::with_tracing::{conv2d, Conv2d};
 use candle::{Module, Result, Tensor, D};
 use candle_nn as nn;
 

candle-transformers/src/models/stable_diffusion/utils.rs

@@ -1,5 +1,4 @@
 use candle::{Device, Result, Tensor};
-use candle_nn::Module;
 
 pub fn linspace(start: f64, stop: f64, steps: usize) -> Result<Tensor> {
     if steps < 1 {
@@ -11,29 +10,3 @@ pub fn linspace(start: f64, stop: f64, steps: usize) -> Result<Tensor> {
         .collect::<Vec<_>>();
     Tensor::from_vec(vs, steps, &Device::Cpu)
 }
-
-// Wrap the conv2d op to provide some tracing.
-#[derive(Debug)]
-pub struct Conv2d {
-    inner: candle_nn::Conv2d,
-    span: tracing::Span,
-}
-
-impl Conv2d {
-    pub fn forward(&self, x: &Tensor) -> Result<Tensor> {
-        let _enter = self.span.enter();
-        self.inner.forward(x)
-    }
-}
-
-pub fn conv2d(
-    in_channels: usize,
-    out_channels: usize,
-    kernel_size: usize,
-    cfg: candle_nn::Conv2dConfig,
-    vs: candle_nn::VarBuilder,
-) -> Result<Conv2d> {
-    let span = tracing::span!(tracing::Level::TRACE, "conv2d");
-    let inner = candle_nn::conv2d(in_channels, out_channels, kernel_size, cfg, vs)?;
-    Ok(Conv2d { inner, span })
-}

candle-transformers/src/models/t5.rs

@@ -1,57 +1,12 @@
 // T5 Text Model
 // https://github.com/huggingface/transformers/blob/main/src/transformers/models/t5/modeling_t5.py
 
+use crate::models::with_tracing::{linear_no_bias, Embedding, Linear};
 use candle::{DType, Device, Module, Result, Tensor, D};
 use candle_nn::{Activation, VarBuilder};
 use serde::Deserialize;
 use std::sync::Arc;
 
-#[derive(Debug)]
-struct Embedding {
-    inner: candle_nn::Embedding,
-    span: tracing::Span,
-}
-
-impl Embedding {
-    fn new(d1: usize, d2: usize, vb: VarBuilder) -> Result<Self> {
-        let inner = candle_nn::embedding(d1, d2, vb)?;
-        let span = tracing::span!(tracing::Level::TRACE, "embedding");
-        Ok(Self { inner, span })
-    }
-
-    fn embeddings(&self) -> &Tensor {
-        self.inner.embeddings()
-    }
-}
-
-impl Module for Embedding {
-    fn forward(&self, xs: &Tensor) -> Result<Tensor> {
-        let _enter = self.span.enter();
-        self.inner.forward(xs)
-    }
-}
-
-#[derive(Debug)]
-struct Linear {
-    inner: candle_nn::Linear,
-    span: tracing::Span,
-}
-
-impl Linear {
-    fn new(d1: usize, d2: usize, vb: VarBuilder) -> Result<Self> {
-        let inner = candle_nn::linear_no_bias(d1, d2, vb)?;
-        let span = tracing::span!(tracing::Level::TRACE, "linear");
-        Ok(Self { inner, span })
-    }
-}
-
-impl Module for Linear {
-    fn forward(&self, xs: &Tensor) -> Result<Tensor> {
-        let _enter = self.span.enter();
-        self.inner.forward(xs)
-    }
-}
-
 fn default_relative_attention_max_distance() -> usize {
     128
 }
@@ -205,8 +160,8 @@ struct T5DenseActDense {
 
 impl T5DenseActDense {
     fn load(vb: VarBuilder, cfg: &Config) -> Result<Self> {
-        let wi = Linear::new(cfg.d_model, cfg.d_ff, vb.pp("wi"))?;
-        let wo = Linear::new(cfg.d_ff, cfg.d_model, vb.pp("wo"))?;
+        let wi = linear_no_bias(cfg.d_model, cfg.d_ff, vb.pp("wi"))?;
+        let wo = linear_no_bias(cfg.d_ff, cfg.d_model, vb.pp("wo"))?;
         Ok(Self {
             wi,
             wo,
@@ -237,9 +192,9 @@ struct T5DenseGatedActDense {
 
 impl T5DenseGatedActDense {
     fn load(vb: VarBuilder, cfg: &Config) -> Result<Self> {
-        let wi_0 = Linear::new(cfg.d_model, cfg.d_ff, vb.pp("wi_0"))?;
-        let wi_1 = Linear::new(cfg.d_model, cfg.d_ff, vb.pp("wi_1"))?;
-        let wo = Linear::new(cfg.d_ff, cfg.d_model, vb.pp("wo"))?;
+        let wi_0 = linear_no_bias(cfg.d_model, cfg.d_ff, vb.pp("wi_0"))?;
+        let wi_1 = linear_no_bias(cfg.d_model, cfg.d_ff, vb.pp("wi_1"))?;
+        let wo = linear_no_bias(cfg.d_ff, cfg.d_model, vb.pp("wo"))?;
         Ok(Self {
             wi_0,
             wi_1,
@@ -334,10 +289,10 @@ impl T5Attention {
         cfg: &Config,
     ) -> Result<Self> {
         let inner_dim = cfg.num_heads * cfg.d_kv;
-        let q = Linear::new(cfg.d_model, inner_dim, vb.pp("q"))?;
-        let k = Linear::new(cfg.d_model, inner_dim, vb.pp("k"))?;
-        let v = Linear::new(cfg.d_model, inner_dim, vb.pp("v"))?;
-        let o = Linear::new(inner_dim, cfg.d_model, vb.pp("o"))?;
+        let q = linear_no_bias(cfg.d_model, inner_dim, vb.pp("q"))?;
+        let k = linear_no_bias(cfg.d_model, inner_dim, vb.pp("k"))?;
+        let v = linear_no_bias(cfg.d_model, inner_dim, vb.pp("v"))?;
+        let o = linear_no_bias(inner_dim, cfg.d_model, vb.pp("o"))?;
         let relative_attention_bias = if has_relative_attention_bias {
             let emb = Embedding::new(
                 cfg.relative_attention_num_buckets,
@@ -772,7 +727,11 @@ impl T5ForConditionalGeneration {
         let lm_head = if tie_word_embeddings {
             None
         } else {
-            Some(Linear::new(cfg.d_model, cfg.vocab_size, vb.pp("lm_head"))?)
+            Some(linear_no_bias(
+                cfg.d_model,
+                cfg.vocab_size,
+                vb.pp("lm_head"),
+            )?)
         };
 
         Ok(Self {