i'm just interested in actually running a prediction with the serving engine and all
i'm also not sure what this is-H "Content-Type: application/octet-stream" -H' NV-InferRequest:batch_size: 1 input { name: "dense_input" dims: [-1, 784] } output { name: "activation_2" cls { count: 1 } }'
Hmm I seems to fit the code 1x784 with float32, no?
platform: "tensorflow_savedmodel" input [ { name: "dense_input" data_type: TYPE_FP32 dims: [-1, 784] } ] output [ { name: "activation_2" data_type: TYPE_FP32 dims: [-1, 10] } ]
i can't get some sort of response from curl
it's from the github issue you sent me but i don't know what the "application" part is or the "NV-InferRequest:...."
well, i have run the keras mnist example that is in the clearml-serving READme. Now I'm just trying to send a request to make a prediction via curl
img.tofile('7.binaryimage',format='binary')
can this be the issue?
ElegantCoyote26 , Hi 🙂
Can you provide an example of what you're trying to do?
fp32 seems to be floating point 32 so my preprocessing seems wrong
So far I have taken one mnist image, and done the following:
` from PIL import Image
import numpy as np
def preprocess(img, format, dtype, h, w, scaling):
sample_img = img.convert('L')
resized_img = sample_img.resize((1, w*h), Image.BILINEAR)
resized = np.array(resized_img)
resized = resized.astype(dtype)
return resized
png img file
img = Image.open('./7.png')
preprocessed img, FP32 formated numpy array
img = preprocess(img, format, "float32", 28, 28, None)
to bytes
img = img.tofile('7.binaryimage',format='binary') curl -X POST 192.168.34.174:8000/v2/models/keras_mnist/versions/1 -H "Content-Type: application/octet-stream" -H' NV-InferRequest:batch_size: 1 input { name: "dense_input" dims: [-1, 784] } output { name: "activation_2" cls { count: 1 } }' --data-binary "@7.binaryimage" `
ElegantCoyote26 what is the model input layer definition? This implies the data format to pass to the serve endpoint
i'm probably sending the request all wrong + i'm not sure how the model expects the input