Getting Error while feeding the input to the onnx model....(RuntimeError: Could not create tensor from given input list)

[3rang]

Hello all,

i'm new to the onnx and onnxruntime. I have been trying to run onnx model with my inputs,but end-up with this error RuntimeError: Could not create tensor from given input list

apart from that, is there any other way to validate onnx model ?? (i'm getting None by using below command)

print(onnx.checker.check_model(modell))

can anyone help me out with any idea....

import onnx
import onnxruntime as rt
import numpy as np
import tensorflow as tf

modell = onnx.load("model1.onnx")

sess = rt.InferenceSession("model1.onnx")

print("----------------------INPUT----------------------------------------")

print("----------------------------------------------")

input_name0 = sess.get_inputs()[0].name
print(input_name0)

input_shape0 = sess.get_inputs()[0].shape
print(input_shape0)

input_type0 = sess.get_inputs()[0].type
print(input_type0)

print("----------------------------------------------")

input_name1 = sess.get_inputs()[1].name
print(input_name1)

input_shape1 = sess.get_inputs()[1].shape
print(input_shape1)

input_type1 = sess.get_inputs()[1].type
print(input_type1)

print("----------------------------------------------")

input_name2 = sess.get_inputs()[2].name
print(input_name2)

input_shape2 = sess.get_inputs()[2].shape
print(input_shape2)

input_type2 = sess.get_inputs()[2].type
print(input_type2)

print("----------------------------------------------")
print(onnx.checker.check_model(modell))

print("----------------------output----------------------------------------")

output_name0 = sess.get_outputs()[0].name
print(output_name0)

output_shape0 = sess.get_outputs()[0].shape
print(output_shape0)

output_type0 = sess.get_outputs()[0].type
print(output_type0)

print("----------------------------------------------")

output_name1 = sess.get_outputs()[1].name
print(output_name1)

output_shape1 = sess.get_outputs()[1].shape
print(output_shape1)

output_type1 = sess.get_outputs()[1].type
print(output_type1)

print("----------------------------------------------")

output_name2 = sess.get_outputs()[2].name
print(output_name2)

output_shape2 = sess.get_outputs()[2].shape
print(output_shape2)

output_type2 = sess.get_outputs()[2].type
print(output_type2)

print("----------------------------------------------")



result = sess.run([output_name0], {input_name0: [0, 9] ,input_name1:[1], input_name2 :[1]})
print(result)

OUTPUT

----------------------INPUT----------------------------------------
----------------------------------------------
input_1:0
['unk__2793', 9]
tensor(int32)
----------------------------------------------
input_2:0
['unk__2794']
tensor(int32)
----------------------------------------------
input_3:0
['unk__2795']
tensor(int32)
----------------------------------------------
Traceback (most recent call last):
  File "/lhome/tarpate/PycharmProjects/pythonProject/main.py", line 94, in <module>
    result = sess.run([output_name0], {input_name0: [0, 9] ,input_name1:[1], input_name2 :[1]})
  File "/lhome/tarpate/PycharmProjects/pythonProject/venv/lib/python3.6/site-packages/onnxruntime/capi/onnxruntime_inference_collection.py", line 188, in run
    return self._sess.run(output_names, input_feed, run_options)
RuntimeError: Could not create tensor from given input list
None
----------------------output----------------------------------------
output_1
['unk__2796', 'unk__2797', 80]
tensor(float)
----------------------------------------------
output_2
['unk__2798', 'unk__2799', 80]
tensor(float)
----------------------------------------------
output_3
['unk__2800', 'unk__2801']
tensor(float)
----------------------------------------------