I have deployed a llama2 model using the model.py file in this repository (the config.pbtxt is the same as this repo). When I send the request to the model the following error appears:
{
"error": "Failed to process the request(s) for model instance 'llama2_0_0', message: AttributeError: 'int' object has no attribute 'decode'\n\nAt:\n /llama2-70b-instruct/llama2/llama2/1/model.py(45): \n /llama2-70b-instruct/llama2/llama2/1/model.py(45): execute\n"
}
it seems that the request can not be resolved correctly in inputs = inputs.as_numpy() in execute function. Would you please propose a Solution?
Hi,
I have deployed a llama2 model using the model.py file in this repository (the
config.pbtxtis the same as this repo). When I send the request to the model the following error appears:{ "error": "Failed to process the request(s) for model instance 'llama2_0_0', message: AttributeError: 'int' object has no attribute 'decode'\n\nAt:\n /llama2-70b-instruct/llama2/llama2/1/model.py(45):\n /llama2-70b-instruct/llama2/llama2/1/model.py(45): execute\n"
}
it seems that the request can not be resolved correctly in
inputs = inputs.as_numpy()in execute function. Would you please propose a Solution?