Refine code.

wit/lit_module.py

@@ -59,11 +59,11 @@ class LitModule(pl.LightningModule):
         labels = batch["labels"][..., 1:]
         labels = labels.contiguous().view(-1)
         label_mask = labels < self.vocab_size
-        logits = logits[label_mask]
+        logits_m = logits[label_mask]
-        labels = labels[label_mask]
+        labels_m = labels[label_mask]
         # m = torch.max(logits, 1).indices.cpu().numpy()
         # ll = labels.cpu().numpy()
-        self.metric_accuracy.update(logits, labels)
+        self.metric_accuracy.update(logits_m, labels_m)
         self.metric_loss.update(loss)
 
     def on_validation_epoch_end(self) -> None:
@@ -71,18 +71,6 @@ class LitModule(pl.LightningModule):
         self.log("accuracy", self.metric_accuracy, rank_zero_only=True)
 
     def configure_optimizers(self):
-        strategy = self.trainer.strategy
-        if isinstance(strategy, pl.strategies.DeepSpeedStrategy):
-            assert "optimizer" not in strategy.config
-            zero_config = strategy.config.get("zero_optimization")
-            if zero_config is not None:
-                if "offload_optimizer" in zero_config:
-                    import deepspeed
-
-                    optimizer = deepspeed.ops.adam.DeepSpeedCPUAdam(
-                        self.trainer.model.parameters(), lr=self.learning_rate
-                    )
-                    return optimizer
         optimizer = torch.optim.AdamW(self.trainer.model.parameters(), lr=self.learning_rate)
         return optimizer
 

wit/train.py

@@ -35,19 +35,26 @@ vocab_size = 4096
 
 
 class SpecialDataset(Dataset):
-    def __init__(self, start=1, end=320, size=32768):  # 1048576
+    def __init__(self, start=1, end=16, size=32768):  # 1048576 32768
         self.size = size
         self.features = []
         a = torch.randint(start, end, [size])
         b = torch.randint(start, end, [size])
         c = torch.randint(start, end, [size])
         d = torch.randint(start, end, [size])
-        # self.data = torch.stack([a, b, a + b, a + b]).permute(1, 0)
+        z = torch.zeros([size]).long()
-        self.data = torch.stack([a, a, a + a]).permute(1, 0)
+        # self.data = torch.stack([a, b, a + b, a + b, a + b * 2]).permute(1, 0)
+        # self.data = torch.stack([a, b, a, a + b / 4]).permute(1, 0).long()
+        # self.data = torch.stack([a, a + 1, a + 2]).permute(1, 0).long()
+        self.data = torch.stack([a, b, a]).permute(1, 0).long()
+        # self.data = torch.stack([a, b, a, a + a / 8, a + a / 4, a + a / 2, a + a]).permute(1, 0).long()
 
         # a = torch.randint(start, end, [size])
         # self.data = torch.stack([a, a, a + a]).permute(1, 0)  # accuracy=0.5
         # self.data = torch.stack([a, a + a, a]).permute(1, 0)  # accuracy=1
+        # 只能有一种算法，而且第一个值不能用于训练
+        # 太陡峭的过度导致难以拟合
+        # 搜索空间太大，难以拟合
 
     def __len__(self):
         return self.size