From fdc8c657b31c24526ac730e2afbcb5ea39f58b5f Mon Sep 17 00:00:00 2001
From: Colin <colin>
Date: Tue, 5 Mar 2024 19:30:15 +0800
Subject: [PATCH] Add accurancy in loss.

---
 test/loss.py | 43 +++++++++++++++++++++++++++++++++----------
 1 file changed, 33 insertions(+), 10 deletions(-)

diff --git a/test/loss.py b/test/loss.py
index a5c5f78..8df1d6a 100644
--- a/test/loss.py
+++ b/test/loss.py
@@ -4,19 +4,42 @@ import torch.nn.functional as F
 import torch.utils.checkpoint
 from torch.nn import CrossEntropyLoss
 import math
+import torchmetrics
 
-shift_logits = torch.zeros((16, 4096))
-shift_logits[:, 2] = 10.0
-shift_labels = (torch.ones(16) * 2).long()
-loss = CrossEntropyLoss()(shift_logits, shift_labels)
-print(loss)
+# shift_logits = torch.zeros((16, 4096))
+# shift_logits[:, 2] = 10.0
+# shift_labels = (torch.ones(16) * 2).long()
+# loss = CrossEntropyLoss()(shift_logits, shift_labels)
+# print(loss)
 
 
-loss = nn.CrossEntropyLoss()
-input = torch.tensor([[1.0, 2.0, 3.0]])
-target = torch.tensor([0]).long()
-output = loss(input, target)
-print(output)
+# loss = nn.CrossEntropyLoss()
+# input = torch.tensor([[1.0, 2.0, 3.0]])
+# target = torch.tensor([0]).long()
+# output = loss(input, target)
+# print(output)
+
+
+target = torch.tensor([0, 1, 2])
+preds = torch.tensor([[0.1, 0.9, 0], [0.3, 10.1, 0.6], [0.2, 0.3, 0.9]])
+accuracy = torchmetrics.Accuracy(task="multiclass", num_classes=3)
+accur = accuracy(preds, target)
+
+
+metric_accuracy = torchmetrics.Accuracy(
+    task="multiclass",
+    num_classes=4096,
+)
+shift_logits = torch.zeros((16, 2, 4096))
+shift_logits[:8, :, 2] = 10.0
+shift_labels = (torch.ones((16, 2)) * 2).long()
+
+label_mask = shift_labels != 4096
+shift_logits = shift_logits[label_mask]
+shift_labels = shift_labels[label_mask]
+
+accur = metric_accuracy(shift_logits, shift_labels)
+metric_accuracy.update(shift_logits, shift_labels)
 
 
 # torch.manual_seed(32)