fixed kwt skip connections

speechcommands/README.md

@@ -40,29 +40,29 @@ python main.py --help
 
 ## Results
 
-After training with the `kwt1` architecture for 100 epochs, you
+After training with the `kwt1` architecture for 10 epochs, you
 should see the following results:
 
 ```
-Epoch: 99 | avg. Train loss 0.581 | avg. Train acc 0.826 | Throughput: 677.37 samples/sec
+Epoch: 9 | avg. Train loss 0.519 | avg. Train acc 0.857 | Throughput: 661.28 samples/sec
-Epoch: 99 | Val acc 0.710
+Epoch: 9 | Val acc 0.861 | Throughput: 2976.54 samples/sec
-Testing best model from Epoch 98
+Testing best model from epoch 9
-Test acc -> 0.687
+Test acc -> 0.841
 ```
 
 For the `kwt2` model, you should see:
 ```
-Epoch: 99 | avg. Train loss 0.137 | avg. Train acc 0.956 | Throughput: 401.47 samples/sec
+Epoch: 9 | avg. Train loss 0.374 | avg. Train acc 0.895 | Throughput: 395.26 samples/sec
-Epoch: 99 | Val acc 0.739
+Epoch: 9 | Val acc 0.879 | Throughput: 1542.44 samples/sec
-Testing best model from Epoch 97
+Testing best model from epoch 9
-Test acc -> 0.718
+Test acc -> 0.861
 ```
 
 Note that this was run on an M1 Macbook Pro with 16GB RAM.
 
 At the time of writing, `mlx` doesn't have built-in `cosine` learning rate
 schedules, which is used along with the AdamW optimizer in the official
-implementaiton. We intend to update this example once these features are added,
+implementation. We intend to update this example once these features are added,
 as well as with appropriate data augmentations.
 
 [^1]: Based one the paper [Keyword Transformer: A Self-Attention Model for Keyword Spotting](https://www.isca-speech.org/archive/interspeech_2021/berg21_interspeech.html)

speechcommands/kwt.py

@@ -47,10 +47,8 @@ class Block(nn.Module):
         self.norm2 = nn.LayerNorm(dim)
 
     def __call__(self, x):
-        x = self.attn(x)
+        x = self.norm1(self.attn(x)) + x
-        x = self.norm1(x)
+        x = self.norm2(self.ff(x)) + x
-        x = self.ff(x)
-        x = self.norm2(x)
         return x
 
 

speechcommands/main.py

@@ -93,7 +93,6 @@ def train_epoch(model, train_iter, optimizer, epoch):
                     )
                 )
             )
-        break
 
     mean_tr_loss = mx.mean(mx.array(losses))
     mean_tr_acc = mx.mean(mx.array(accs))
@@ -104,13 +103,18 @@ def train_epoch(model, train_iter, optimizer, epoch):
 def test_epoch(model, test_iter):
     model.train(False)
     accs = []
+    throughput = []
     for batch_counter, batch in enumerate(test_iter):
         x = mx.array(batch["audio"])
         y = mx.array(batch["label"])
+        tic = time.perf_counter()
         acc = eval_fn(model, x, y)
         accs.append(acc.item())
+        toc = time.perf_counter()
+        throughput.append(x.shape[0] / (toc - tic))
     mean_acc = mx.mean(mx.array(accs))
-    return mean_acc
+    mean_throughput = mx.mean(mx.array(throughput))
+    return mean_acc, mean_throughput
 
 
 def main(args):
@@ -141,8 +145,8 @@ def main(args):
             )
         )
 
-        val_acc = test_epoch(model, val_data)
+        val_acc, val_throughput = test_epoch(model, val_data)
-        print(f"Epoch: {epoch} | Val acc {val_acc.item():.3f}")
+        print(f"Epoch: {epoch} | Val acc {val_acc.item():.3f} | Throughput: {val_throughput.item():.2f} samples/sec")
 
         if val_acc >= best_acc:
             best_acc = val_acc
@@ -151,7 +155,7 @@ def main(args):
     print(f"Testing best model from epoch {best_epoch}")
     model.update(best_params)
     test_data = prepare_dataset(args.batch_size, "test")
-    test_acc = test_epoch(model, test_data)
+    test_acc, _ = test_epoch(model, test_data)
     print(f"Test acc -> {test_acc.item():.3f}")