kserve/docs/samples/v1beta1/torchserve/v1/bert/README.md

TorchServe example with Huggingface bert model

In this example we will show how to serve Huggingface Transformers with TorchServe on KServe.

Model archive file creation

Clone pytorch/serve repository, navigate to examples/Huggingface_Transformers and follow the steps for creating the MAR file including serialized model and other dependent files. TorchServe supports both eager model and torchscript and here we save as the pretrained model.

torch-model-archiver --model-name BERTSeqClassification --version 1.0 \
--serialized-file Transformer_model/pytorch_model.bin \
--handler ./Transformer_handler_generalized.py \
--extra-files "Transformer_model/config.json,./setup_config.json,./Seq_classification_artifacts/index_to_name.json"

Create the InferenceService

Apply the CRD

kubectl apply -f bert.yaml

Expected Output

$inferenceservice.serving.kubeflow.org/torchserve-bert created

Run a prediction

The first step is to determine the ingress IP and ports and set INGRESS_HOST and INGRESS_PORT

MODEL_NAME=torchserve-bert
SERVICE_HOSTNAME=$(kubectl get inferenceservice ${MODEL_NAME} -n <namespace> -o jsonpath='{.status.url}' | cut -d "/" -f 3)

curl -v -H "Host: ${SERVICE_HOSTNAME}" http://${INGRESS_HOST}:${INGRESS_PORT}/v1/models/BERTSeqClassification:predict -d ./sample_text.txt

Expected Output

*   Trying 44.239.20.204...
* Connected to a881f5a8c676a41edbccdb0a394a80d6-2069247558.us-west-2.elb.amazonaws.com (44.239.20.204) port 80 (#0)
> PUT /v1/models/BERTSeqClassification:predict HTTP/1.1
> Host: torchserve-bert.kserve-test.example.com
> User-Agent: curl/7.47.0
> Accept: */*
> Content-Length: 79
> Expect: 100-continue
>
< HTTP/1.1 100 Continue
* We are completely uploaded and fine
< HTTP/1.1 200 OK
< cache-control: no-cache; no-store, must-revalidate, private
< content-length: 8
< date: Wed, 04 Nov 2020 10:54:49 GMT
< expires: Thu, 01 Jan 1970 00:00:00 UTC
< pragma: no-cache
< x-request-id: 4b54d3ac-185f-444c-b344-b8a785fdeb50
< x-envoy-upstream-service-time: 2085
< server: istio-envoy
<
* Connection #0 to host torchserve-bert.kserve-test.example.com left intact
Accepted

Captum Explanations

In order to understand the word importances and attributions when we make an explanation Request, we use Captum Insights for the HuggingFace Transformers pre-trained model.

MODEL_NAME=torchserve-bert
SERVICE_HOSTNAME=$(kubectl get inferenceservice ${MODEL_NAME} -n <namespace> -o jsonpath='{.status.url}' | cut -d "/" -f 3)

curl -v -H "Host: ${SERVICE_HOSTNAME}" http://${INGRESS_HOST}:${INGRESS_PORT}/v1/models/BERTSeqClassification:explain -d ./sample_text.txt

Expected output

*   Trying ::1:8080...
* Connected to localhost (::1) port 8080 (#0)
> POST /v1/models/BERTSeqClassification:explain HTTP/1.1
> Host: torchserve-bert.default.example.com
> User-Agent: curl/7.73.0
> Accept: */*
> Content-Length: 84
> Content-Type: application/x-www-form-urlencoded
>Handling connection for 8080
 
* upload completely sent off: 84 out of 84 bytes
* Mark bundle as not supporting multiuse
< HTTP/1.1 200 OK
< content-length: 292
< content-type: application/json; charset=UTF-8
< date: Sun, 27 Dec 2020 05:53:52 GMT
< server: istio-envoy
< x-envoy-upstream-service-time: 5769
< 
* Connection #0 to host localhost left intact
{"explanations": [{"importances": [0.0, -0.6324463574494716, -0.033115653530477414, 0.2681695752722339, -0.29124745608778546, 0.5422589681903883, -0.3848768219546909, 0.0], 
"words": ["[CLS]", "bloomberg", "has", "reported", "on", "the", "economy", "[SEP]"], "delta": -0.0007350619859377225}]}