我想让我的模型使用自生成的文本图片来检测角度（360个类别）。为了获得更多训练信息，训练集图片在每个epoch都会重新生成并随机旋转。然而，模型似乎没有在学习，因为它对每张图片都预测出相同的类别。我已经尝试更改批量大小、优化器、学习率、使用更复杂的模型，但这些都没有帮助解决问题。
在这个例子中，我使用了500个训练样本，50个验证样本和10个测试样本。我还尝试过使用高达2000个训练样本，但同样的问题仍然存在。
这是我的输出：
Using TensorFlow backend.WARNING:tensorflow:From /home/lisa/.local/lib/python3.7/site-packages/keras/backend/tensorflow_backend.py:4070: The name tf.nn.max_pool is deprecated. Please use tf.nn.max_pool2d instead.Model: "model_1"_________________________________________________________________Layer (type) Output Shape Param # =================================================================input_1 (InputLayer) (None, 224, 224, 3) 0 _________________________________________________________________conv2d_1 (Conv2D) (None, 222, 222, 32) 896 _________________________________________________________________max_pooling2d_1 (MaxPooling2 (None, 111, 111, 32) 0 _________________________________________________________________conv2d_2 (Conv2D) (None, 109, 109, 64) 18496 _________________________________________________________________max_pooling2d_2 (MaxPooling2 (None, 54, 54, 64) 0 _________________________________________________________________conv2d_3 (Conv2D) (None, 52, 52, 128) 73856 _________________________________________________________________max_pooling2d_3 (MaxPooling2 (None, 26, 26, 128) 0 _________________________________________________________________conv2d_4 (Conv2D) (None, 24, 24, 128) 147584 _________________________________________________________________max_pooling2d_4 (MaxPooling2 (None, 12, 12, 128) 0 _________________________________________________________________flatten_1 (Flatten) (None, 18432) 0 _________________________________________________________________dense_1 (Dense) (None, 512) 9437696 _________________________________________________________________dense_2 (Dense) (None, 360) 184680 =================================================================Total params: 9,863,208Trainable params: 9,863,208Non-trainable params: 0_________________________________________________________________2019-11-06 11:08:47.885295: I tensorflow/core/platform/cpu_feature_guard.cc:142] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA2019-11-06 11:08:47.901431: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 3408000000 Hz2019-11-06 11:08:47.902091: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x55f4487aac50 executing computations on platform Host. Devices:2019-11-06 11:08:47.902139: I tensorflow/compiler/xla/service/service.cc:175] StreamExecutor device (0): <undefined>, <undefined>2019-11-06 11:08:47.903354: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcuda.so.12019-11-06 11:08:47.921001: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero2019-11-06 11:08:47.921953: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1640] Found device 0 with properties: name: GeForce GTX 970 major: 5 minor: 2 memoryClockRate(GHz): 1.1775pciBusID: 0000:01:00.02019-11-06 11:08:47.922112: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudart.so.10.02019-11-06 11:08:47.922988: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcublas.so.10.02019-11-06 11:08:47.923739: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcufft.so.10.02019-11-06 11:08:47.923921: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcurand.so.10.02019-11-06 11:08:47.924921: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcusolver.so.10.02019-11-06 11:08:47.925684: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcusparse.so.10.02019-11-06 11:08:47.928111: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudnn.so.72019-11-06 11:08:47.928199: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero2019-11-06 11:08:47.929103: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero2019-11-06 11:08:47.929818: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1763] Adding visible gpu devices: 02019-11-06 11:08:47.929844: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudart.so.10.02019-11-06 11:08:47.976192: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1181] Device interconnect StreamExecutor with strength 1 edge matrix:2019-11-06 11:08:47.976213: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1187] 0 2019-11-06 11:08:47.976219: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1200] 0: N 2019-11-06 11:08:47.976372: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero2019-11-06 11:08:47.977217: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero2019-11-06 11:08:47.978039: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:1005] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero2019-11-06 11:08:47.978851: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1326] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 3466 MB memory) -> physical GPU (device: 0, name: GeForce GTX 970, pci bus id: 0000:01:00.0, compute capability: 5.2)2019-11-06 11:08:47.980313: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x55f449158000 executing computations on platform CUDA. Devices:2019-11-06 11:08:47.980326: I tensorflow/compiler/xla/service/service.cc:175] StreamExecutor device (0): GeForce GTX 970, Compute Capability 5.2WARNING:tensorflow:From /home/lisa/.local/lib/python3.7/site-packages/keras/backend/tensorflow_backend.py:422: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.Epoch 1/502019-11-06 11:08:48.922378: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcublas.so.10.02019-11-06 11:08:49.080712: I tensorflow/stream_executor/platform/default/dso_loader.cc:42] Successfully opened dynamic library libcudnn.so.716/16 [==============================] - 3s 199ms/step - loss: 10271548.3852 - mse_angle: 88.4758 - val_loss: 6.0310 - val_mse_angle: 83.5972Epoch 2/5016/16 [==============================] - 1s 84ms/step - loss: 6.0294 - mse_angle: 87.3988 - val_loss: 6.2498 - val_mse_angle: 90.8889Epoch 3/5016/16 [==============================] - 1s 82ms/step - loss: 6.9000 - mse_angle: 90.9215 - val_loss: 6.2606 - val_mse_angle: 96.1042Epoch 4/5016/16 [==============================] - 1s 82ms/step - loss: 6.0261 - mse_angle: 90.2238 - val_loss: 6.1281 - val_mse_angle: 89.1111Epoch 5/5016/16 [==============================] - 1s 82ms/step - loss: 6.0339 - mse_angle: 90.6246 - val_loss: 6.1609 - val_mse_angle: 84.5764Epoch 6/5016/16 [==============================] - 1s 82ms/step - loss: 5.9953 - mse_angle: 90.6105 - val_loss: 6.0373 - val_mse_angle: 97.3819Epoch 7/5016/16 [==============================] - 1s 82ms/step - loss: 5.9419 - mse_angle: 90.0617 - val_loss: 6.0082 - val_mse_angle: 99.2257Epoch 8/5016/16 [==============================] - 1s 82ms/step - loss: 5.9563 - mse_angle: 89.2258 - val_loss: 6.0243 - val_mse_angle: 99.2257Epoch 9/5016/16 [==============================] - 1s 83ms/step - loss: 5.9515 - mse_angle: 92.9902 - val_loss: 6.0726 - val_mse_angle: 87.7812Epoch 10/5016/16 [==============================] - 1s 82ms/step - loss: 5.9554 - mse_angle: 89.0434 - val_loss: 6.0980 - val_mse_angle: 81.9757Epoch 11/5016/16 [==============================] - 1s 82ms/step - loss: 5.9761 - mse_angle: 90.9699 - val_loss: 6.1573 - val_mse_angle: 99.1910Epoch 12/5016/16 [==============================] - 1s 83ms/step - loss: 5.9674 - mse_angle: 87.5254 - val_loss: 6.1502 - val_mse_angle: 91.5312Epoch 13/5016/16 [==============================] - 1s 82ms/step - loss: 5.9457 - mse_angle: 90.9098 - val_loss: 6.1447 - val_mse_angle: 89.7708Epoch 14/5016/16 [==============================] - 1s 82ms/step - loss: 5.9803 - mse_angle: 92.3281 - val_loss: 6.1520 - val_mse_angle: 97.5417Epoch 15/5016/16 [==============================] - 1s 82ms/step - loss: 5.9663 - mse_angle: 91.3766 - val_loss: 6.1332 - val_mse_angle: 81.1562Epoch 16/5016/16 [==============================] - 1s 82ms/step - loss: 5.9707 - mse_angle: 89.2891 - val_loss: 6.0442 - val_mse_angle: 88.7361Epoch 17/5016/16 [==============================] - 1s 82ms/step - loss: 5.9691 - mse_angle: 87.9980 - val_loss: 5.8971 - val_mse_angle: 81.1562Epoch 18/5016/16 [==============================] - 1s 82ms/step - loss: 5.9675 - mse_angle: 87.8605 - val_loss: 5.9070 - val_mse_angle: 81.1562Epoch 19/5016/16 [==============================] - 1s 81ms/step - loss: 5.9816 - mse_angle: 88.3820 - val_loss: 6.0384 - val_mse_angle: 90.0694Epoch 20/5016/16 [==============================] - 1s 82ms/step - loss: 6.0144 - mse_angle: 91.3855 - val_loss: 6.1066 - val_mse_angle: 90.0694Epoch 21/5016/16 [==============================] - 1s 82ms/step - loss: 5.9556 - mse_angle: 92.5727 - val_loss: 6.2307 - val_mse_angle: 86.2465Epoch 22/5016/16 [==============================] - 1s 83ms/step - loss: 5.9522 - mse_angle: 90.1418 - val_loss: 6.1750 - val_mse_angle: 81.9062Epoch 23/5016/16 [==============================] - 1s 81ms/step - loss: 5.9603 - mse_angle: 88.3703 - val_loss: 6.0286 - val_mse_angle: 81.9062Epoch 24/5016/16 [==============================] - 1s 82ms/step - loss: 5.9608 - mse_angle: 90.1531 - val_loss: 5.9816 - val_mse_angle: 97.9549Epoch 25/5016/16 [==============================] - 1s 82ms/step - loss: 5.9764 - mse_angle: 88.8660 - val_loss: 6.0606 - val_mse_angle: 89.0174Epoch 26/5016/16 [==============================] - 1s 82ms/step - loss: 5.9771 - mse_angle: 90.2336 - val_loss: 6.0759 - val_mse_angle: 83.8507Epoch 27/5016/16 [==============================] - 1s 82ms/step - loss: 6.0073 - mse_angle: 90.3863 - val_loss: 6.0298 - val_mse_angle: 83.8507Epoch 28/5016/16 [==============================] - 1s 82ms/step - loss: 5.9664 - mse_angle: 89.0832 - val_loss: 5.9718 - val_mse_angle: 83.5972Epoch 29/5016/16 [==============================] - 1s 82ms/step - loss: 5.9445 - mse_angle: 88.3340 - val_loss: 5.9844 - val_mse_angle: 82.4306Epoch 30/5016/16 [==============================] - 1s 81ms/step - loss: 5.9596 - mse_angle: 90.2934 - val_loss: 5.8805 - val_mse_angle: 83.0521Epoch 31/5016/16 [==============================] - 1s 83ms/step - loss: 5.9729 - mse_angle: 91.9238 - val_loss: 5.9500 - val_mse_angle: 84.4444Epoch 32/5016/16 [==============================] - 1s 83ms/step - loss: 5.9743 - mse_angle: 90.0250 - val_loss: 6.0221 - val_mse_angle: 97.5556Epoch 33/5016/16 [==============================] - 1s 81ms/step - loss: 5.9469 - mse_angle: 86.5922 - val_loss: 6.0201 - val_mse_angle: 87.6076Epoch 34/5016/16 [==============================] - 1s 82ms/step - loss: 5.9822 - mse_angle: 93.8836 - val_loss: 5.9119 - val_mse_angle: 81.3472Epoch 35/5016/16 [==============================] - 1s 81ms/step - loss: 5.9751 - mse_angle: 88.9707 - val_loss: 5.9052 - val_mse_angle: 99.3993Epoch 36/5016/16 [==============================] - 1s 83ms/step - loss: 5.9564 - mse_angle: 89.6219 - val_loss: 5.9162 - val_mse_angle: 92.5278Epoch 37/5016/16 [==============================] - 1s 82ms/step - loss: 5.9864 - mse_angle: 94.1816 - val_loss: 5.9559 - val_mse_angle: 90.5278Epoch 38/5016/16 [==============================] - 1s 82ms/step - loss: 5.9566 - mse_angle: 88.3102 - val_loss: 6.0087 - val_mse_angle: 99.3993Epoch 39/5016/16 [==============================] - 1s 83ms/step - loss: 5.9639 - mse_angle: 91.0492 - val_loss: 5.9907 - val_mse_angle: 94.2361Epoch 40/5016/16 [==============================] - 1s 83ms/step - loss: 5.9792 - mse_angle: 88.0059 - val_loss: 5.8827 - val_mse_angle: 94.3056Epoch 41/5016/16 [==============================] - 1s 82ms/step - loss: 5.9297 - mse_angle: 92.0566 - val_loss: 5.8013 - val_mse_angle: 94.6319Epoch 42/5016/16 [==============================] - 1s 84ms/step - loss: 5.9666 - mse_angle: 88.4168 - val_loss: 5.8768 - val_mse_angle: 99.4826Epoch 43/5016/16 [==============================] - 1s 82ms/step - loss: 5.9887 - mse_angle: 90.3191 - val_loss: 5.9197 - val_mse_angle: 96.8611Epoch 44/5016/16 [==============================] - 1s 82ms/step - loss: 5.9889 - mse_angle: 87.8867 - val_loss: 5.8738 - val_mse_angle: 96.6875Epoch 45/5016/16 [==============================] - 1s 83ms/step - loss: 5.9694 - mse_angle: 92.4437 - val_loss: 5.8639 - val_mse_angle: 98.7222Epoch 46/5016/16 [==============================] - 1s 82ms/step - loss: 5.9560 - mse_angle: 89.9125 - val_loss: 5.8387 - val_mse_angle: 82.4965Epoch 47/5016/16 [==============================] - 1s 82ms/step - loss: 5.9468 - mse_angle: 89.7066 - val_loss: 5.9525 - val_mse_angle: 87.1632Epoch 48/5016/16 [==============================] - 1s 83ms/step - loss: 6.0111 - mse_angle: 89.5977 - val_loss: 5.9091 - val_mse_angle: 96.6875Epoch 49/5016/16 [==============================] - 1s 82ms/step - loss: 5.9648 - mse_angle: 89.0430 - val_loss: 5.9656 - val_mse_angle: 92.8368Epoch 50/5016/16 [==============================] - 1s 82ms/step - loss: 5.9234 - mse_angle: 91.1891 - val_loss: 5.9717 - val_mse_angle: 99.2257for image 0 angle: 312, pred: 46for image 1 angle: 202, pred: 46for image 2 angle: 235, pred: 46for image 3 angle: 286, pred: 46for image 4 angle: 226, pred: 46for image 5 angle: 76, pred: 46for image 6 angle: 91, pred: 46for image 7 angle: 91, pred: 46for image 8 angle: 97, pred: 46for image 9 angle: 263, pred: 46这是我的model.py文件：
...要运行代码，需要将其放置在一个包含三个空文件夹（val, train, test）的目录中，并准备create_test_images.py文件：
...非常感谢任何建议！
编辑：删除了两行未使用的代码
回答：
正如我在评论中提到的，使用您提供的代码，我能够重现您的问题，并将其从一个具有260个类别的分类问题重新表述为一个回归问题。
因此，我将输出神经元的数量更改为仅一个，并使用sigmoid激活函数，将标签更改为连续数字并进行归一化（除以360）以得到0到1之间的数字，将损失函数更改为MSE，并使用优化器的默认值。
通过这些修改，我在训练10个epoch后获得了以下结果：
for image 0 angle: 0.7416666666666667, pred: [0.7266706]for image 1 angle: 0.8111111111111111, pred: [0.8449749]for image 2 angle: 0.7777777777777778, pred: [0.84269005]for image 3 angle: 0.12222222222222222, pred: [0.14173588]for image 4 angle: 0.7388888888888889, pred: [0.730219]for image 5 angle: 0.9694444444444444, pred: [0.9117564]for image 6 angle: 0.075, pred: [0.07597628]for image 7 angle: 0.29444444444444445, pred: [0.1829494]for image 8 angle: 0.10277777777777777, pred: [0.12209181]for image 9 angle: 0.21388888888888888, pred: [0.31544465]