Blip attention mask + readme (#1146)

* Add the attention mask to the blip model.

* Add a readme.

candle-examples/examples/blip/README.md

@@ -0,0 +1,19 @@
+# candle-blip
+
+The
+[blip-image-captioning](https://huggingface.co/Salesforce/blip-image-captioning-base)
+model can generate captions for an input image.
+
+## Running on an example
+
+```bash
+cargo run --example blip --release -- --image candle-examples/examples/yolo-v8/assets/bike.jpg
+```
+
+```
+Running on CPU, to run on GPU, build this example with `--features cuda`
+loaded image Tensor[dims 3, 384, 384; f32]
+model built
+several cyclists are riding down a road with cars behind them%
+```
+![Leading group, Giro d'Italia 2021](../yolo-v8/assets/bike.jpg)

candle-transformers/src/models/blip_text.rs

@@ -105,7 +105,12 @@ impl TextSelfAttention {
         .permute((0, 2, 1, 3))
     }
 
-    fn forward(&self, xs: &Tensor, encoder_hidden_states: Option<&Tensor>) -> Result<Tensor> {
+    fn forward(
+        &self,
+        xs: &Tensor,
+        encoder_hidden_states: Option<&Tensor>,
+        attention_mask: Option<&Tensor>,
+    ) -> Result<Tensor> {
         let query = self
             .transpose_for_scores(&self.query.forward(xs)?)?
             .contiguous()?;
@@ -127,6 +132,10 @@ impl TextSelfAttention {
         let value = value.contiguous()?;
         let attention_scores = query.matmul(&key.t()?)?;
         let attention_scores = (attention_scores * self.attention_scale)?;
+        let attention_scores = match attention_mask {
+            Some(mask) => attention_scores.broadcast_add(mask)?,
+            None => attention_scores,
+        };
         let attention_probs = candle_nn::ops::softmax_last_dim(&attention_scores)?;
         attention_probs
             .matmul(&value)?
@@ -166,8 +175,15 @@ impl TextAttention {
         Ok(Self { self_, output })
     }
 
-    fn forward(&self, xs: &Tensor, encoder_hidden_states: Option<&Tensor>) -> Result<Tensor> {
+    fn forward(
-        let self_outputs = self.self_.forward(xs, encoder_hidden_states)?;
+        &self,
+        xs: &Tensor,
+        encoder_hidden_states: Option<&Tensor>,
+        attention_mask: Option<&Tensor>,
+    ) -> Result<Tensor> {
+        let self_outputs = self
+            .self_
+            .forward(xs, encoder_hidden_states, attention_mask)?;
         self.output.forward(&self_outputs, xs)
     }
 }
@@ -238,10 +254,15 @@ impl TextLayer {
         })
     }
 
-    fn forward(&self, xs: &Tensor, encoder_hidden_states: &Tensor) -> Result<Tensor> {
+    fn forward(
-        let attention_output = self.attention.forward(xs, None)?;
+        &self,
+        xs: &Tensor,
+        encoder_hidden_states: &Tensor,
+        attention_mask: &Tensor,
+    ) -> Result<Tensor> {
+        let attention_output = self.attention.forward(xs, None, Some(attention_mask))?;
         let attention_output = match &self.cross_attention {
-            Some(ca) => ca.forward(&attention_output, Some(encoder_hidden_states))?,
+            Some(ca) => ca.forward(&attention_output, Some(encoder_hidden_states), None)?,
             None => candle::bail!("expected some cross-attn"),
         };
         let intermediate_output = self.intermediate.forward(&attention_output)?;
@@ -265,10 +286,15 @@ impl TextEncoder {
         Ok(Self { layers })
     }
 
-    fn forward(&self, xs: &Tensor, encoder_hidden_states: &Tensor) -> Result<Tensor> {
+    fn forward(
+        &self,
+        xs: &Tensor,
+        encoder_hidden_states: &Tensor,
+        attention_mask: &Tensor,
+    ) -> Result<Tensor> {
         let mut xs = xs.clone();
         for layer in self.layers.iter() {
-            xs = layer.forward(&xs, encoder_hidden_states)?
+            xs = layer.forward(&xs, encoder_hidden_states, attention_mask)?
         }
         Ok(xs)
     }
@@ -384,11 +410,16 @@ impl TextModel {
         })
     }
 
-    fn forward(&self, input_ids: &Tensor, encoder_hidden_states: &Tensor) -> Result<Tensor> {
+    fn forward(
+        &self,
+        input_ids: &Tensor,
+        encoder_hidden_states: &Tensor,
+        attention_mask: &Tensor,
+    ) -> Result<Tensor> {
         let embedding_output = self.embeddings.forward(input_ids)?;
-        let sequence_output = self
+        let sequence_output =
-            .encoder
+            self.encoder
-            .forward(&embedding_output, encoder_hidden_states)?;
+                .forward(&embedding_output, encoder_hidden_states, attention_mask)?;
         // We're interested in the sequence-output rather than the pooled-output.
         Ok(sequence_output)
     }
@@ -408,7 +439,12 @@ impl TextLMHeadModel {
     }
 
     pub fn forward(&self, input_ids: &Tensor, encoder_hidden_states: &Tensor) -> Result<Tensor> {
-        let sequence_output = self.bert.forward(input_ids, encoder_hidden_states)?;
+        let seq_len = input_ids.dim(1)?;
+        let mask: Vec<_> = (0..seq_len)
+            .flat_map(|i| (0..seq_len).map(move |j| if j > i { f32::NEG_INFINITY } else { 0f32 }))
+            .collect();
+        let mask = Tensor::from_vec(mask, (seq_len, seq_len), input_ids.device())?;
+        let sequence_output = self.bert.forward(input_ids, encoder_hidden_states, &mask)?;
         let prediction_scores = self.cls.forward(&sequence_output)?;
         // return_logits is false so we don't discard the last sequence element.
         Ok(prediction_scores)