elvenson/stable_diffusion_keras_inference 📝✓🔢 → 🖼️

▶️ 102 runs 📅 Apr 2023 ⚙️ Cog 0.7.2 🔗 GitHub 📄 Paper ⚖️ License

image-upscaling nature text-to-image

About

Stable Diffusion inference using Keras model

Example Output

Prompt:

"Teddy bears conducting machine learning research"

Output

Performance Metrics

96.34s Prediction Time

96.46s Total Time

All Input Parameters

{
  "prompt": "Teddy bears conducting machine learning research",
  "version": "2",
  "image_size": "640x960",
  "num_outputs": 1,
  "num_inference_steps": 50,
  "unconditional_guidance_scale": 7.5
}

Input Parameters

prompt (required) Type: string: Image prompt.
upscale Type: booleanDefault: false: Whether to upscale the image by 4 times
num_outputs Type: integerDefault: 1Range: 1 - 4: Number of images to output.
num_inference_steps Type: integerDefault: 50Range: 1 - 500: Number of denoising steps
unconditional_guidance_scale Type: numberDefault: 7.5Range: 1 - 20: Scale for unconditional guidance

Output Schema

Output

Type: array • Items Type: string • Items Format: uri

Example Execution Logs

By using this model checkpoint, you acknowledge that its usage is subject to the terms of the CreativeML Open RAIL++-M license at https://github.com/Stability-AI/stablediffusion/main/LICENSE-MODEL
2023-04-17 10:11:19.537738: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 4 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 320 input_feature_map_count: 4 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.551945: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 320 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 320 input_feature_map_count: 320 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.562262: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 320 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 320 input_feature_map_count: 320 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 2 2  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.571999: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 320 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 320 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.578584: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 320 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 320 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.589934: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 640 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 640 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.600637: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 640 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 640 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 2 2  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.614211: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 640 spatial: 32 20  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 640 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.620868: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 640 spatial: 32 20  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 640 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.640721: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1280 spatial: 32 20  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 1280 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.663433: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1280 spatial: 32 20  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 1280 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 2 2  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.684716: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1280 spatial: 16 10  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 1280 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.721584: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 2560 spatial: 16 10  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 2560 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.733397: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 2560 spatial: 16 10  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 2560 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.753188: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1280 spatial: 32 20  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 1280 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.792489: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 2560 spatial: 32 20  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 2560 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.808472: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 2560 spatial: 32 20  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 2560 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.837135: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1920 spatial: 32 20  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 1920 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.850048: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1920 spatial: 32 20  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 1920 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.874505: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1280 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 1280 input_feature_map_count: 1280 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.902352: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1920 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 1920 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.917671: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1920 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 1920 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.934960: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1280 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 1280 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.946733: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 1280 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 1280 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.960775: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 960 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 960 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.970355: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 960 spatial: 64 40  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 960 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:19.989953: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 640 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 640 input_feature_map_count: 640 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:20.017115: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 960 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 320 input_feature_map_count: 960 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:20.037880: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 960 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 320 input_feature_map_count: 960 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:20.056367: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 640 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 320 input_feature_map_count: 640 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:20.072479: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 640 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 320 input_feature_map_count: 640 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:20.078947: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 320 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 4 input_feature_map_count: 320 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
 1/50 [..............................] - ETA: 40:44
 2/50 [>.............................] - ETA: 12s  
 3/50 [>.............................] - ETA: 12s
 4/50 [=>............................] - ETA: 12s
 5/50 [==>...........................] - ETA: 11s
 6/50 [==>...........................] - ETA: 11s
 7/50 [===>..........................] - ETA: 11s
 8/50 [===>..........................] - ETA: 11s
 9/50 [====>.........................] - ETA: 10s
10/50 [=====>........................] - ETA: 10s
11/50 [=====>........................] - ETA: 10s
12/50 [======>.......................] - ETA: 10s
13/50 [======>.......................] - ETA: 9s 
14/50 [=======>......................] - ETA: 9s
15/50 [========>.....................] - ETA: 9s
16/50 [========>.....................] - ETA: 9s
17/50 [=========>....................] - ETA: 8s
18/50 [=========>....................] - ETA: 8s
19/50 [==========>...................] - ETA: 8s
20/50 [===========>..................] - ETA: 8s
21/50 [===========>..................] - ETA: 7s
22/50 [============>.................] - ETA: 7s
23/50 [============>.................] - ETA: 7s
24/50 [=============>................] - ETA: 6s
25/50 [==============>...............] - ETA: 6s
26/50 [==============>...............] - ETA: 6s
27/50 [===============>..............] - ETA: 6s
28/50 [===============>..............] - ETA: 5s
29/50 [================>.............] - ETA: 5s
30/50 [=================>............] - ETA: 5s
31/50 [=================>............] - ETA: 5s
32/50 [==================>...........] - ETA: 4s
33/50 [==================>...........] - ETA: 4s
34/50 [===================>..........] - ETA: 4s
35/50 [====================>.........] - ETA: 4s
36/50 [====================>.........] - ETA: 3s
37/50 [=====================>........] - ETA: 3s
38/50 [=====================>........] - ETA: 3s
39/50 [======================>.......] - ETA: 2s
40/50 [=======================>......] - ETA: 2s
41/50 [=======================>......] - ETA: 2s
42/50 [========================>.....] - ETA: 2s
43/50 [========================>.....] - ETA: 1s
44/50 [=========================>....] - ETA: 1s
45/50 [==========================>...] - ETA: 1s
46/50 [==========================>...] - ETA: 1s
47/50 [===========================>..] - ETA: 0s
48/50 [===========================>..] - ETA: 0s
49/50 [============================>.] - ETA: 0s
50/50 [==============================] - 63s 267ms/step
2023-04-17 10:11:52.941749: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 4 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 4 input_feature_map_count: 4 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:52.950767: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 4 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 512 input_feature_map_count: 4 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:52.966594: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 512 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 512 input_feature_map_count: 512 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:52.991896: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 512 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 512 input_feature_map_count: 512 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:53.009837: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 512 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 512 input_feature_map_count: 512 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:53.029310: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 512 spatial: 128 80  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 512 input_feature_map_count: 512 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:53.077621: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 512 spatial: 256 160  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 512 input_feature_map_count: 512 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:53.162894: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 512 spatial: 256 160  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 512 input_feature_map_count: 512 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:53.355331: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 512 spatial: 512 320  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 512 input_feature_map_count: 512 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:53.537616: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 512 spatial: 512 320  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 256 input_feature_map_count: 512 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:53.694083: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 512 spatial: 512 320  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 256 input_feature_map_count: 512 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:53.827837: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 256 spatial: 512 320  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 256 input_feature_map_count: 256 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:53.929183: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 256 spatial: 512 320  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 256 input_feature_map_count: 256 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:54.288489: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 256 spatial: 1024 640  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 256 input_feature_map_count: 256 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:54.595427: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 256 spatial: 1024 640  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 128 input_feature_map_count: 256 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:54.883180: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 256 spatial: 1024 640  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 128 input_feature_map_count: 256 layout: OutputInputYX shape: 1 1 }
{zero_padding: 0 0  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:55.145238: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 128 spatial: 1024 640  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 128 input_feature_map_count: 128 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:55.331951: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 128 spatial: 1024 640  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 128 input_feature_map_count: 128 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.
2023-04-17 10:11:55.435626: I tensorflow/compiler/xla/stream_executor/cuda/cuda_dnn.cc:5340] Disabling cuDNN frontend for the following convolution:
input: {count: 1 feature_map_count: 128 spatial: 1024 640  value_min: 0.000000 value_max: 0.000000 layout: BatchDepthYX}
filter: {output_feature_map_count: 3 input_feature_map_count: 128 layout: OutputInputYX shape: 3 3 }
{zero_padding: 1 1  pad_alignment: default filter_strides: 1 1  dilation_rates: 1 1 }
... because it uses an identity activation.

Version Details

Version ID: 9fed3df7cc8aa2eb615f762740f30643d4cd5d324eb807a2bd0e585e4acf13d5
Version Created: June 10, 2023

Run on Replicate →