# Rasa-server | 2023-01-19 07:30:56 ERROR rasa.core.agent - Could not load model due to Error initializing graph component for node run\_LanguageModelFeaturizer1

**URL:** <https://forum.rasa.com/t/rasa-server-2023-01-19-0756-error-rasa-core-agent-could-not-load-model-due-to-error-initializing-graph-component-for-node-run-languagemodelfeaturizer1/56910>\
**Category:** Rasa Open Source\
**Created:** [January 19, 2023, 8:04am UTC](https://forum.rasa.com/t/rasa-server-2023-01-19-0756-error-rasa-core-agent-could-not-load-model-due-to-error-initializing-graph-component-for-node-run-languagemodelfeaturizer1/56910 "2023-01-19T08:04:50Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![gayathrik](https://avatars.discourse-cdn.com/v4/letter/g/958977/32.png) [@gayathrik](https://forum.rasa.com/u/gayathrik)\
**Post date:** [January 19, 2023, 8:04am UTC](https://forum.rasa.com/t/rasa-server-2023-01-19-0756-error-rasa-core-agent-could-not-load-model-due-to-error-initializing-graph-component-for-node-run-languagemodelfeaturizer1/56910/1 "2023-01-19T08:04:50Z")

</div>

hi, I am using ktrain distilbert model for intent classification, i am trying to build custom graph components, from typing import Dict, Text, Any, List import os from typing import Dict, Any, Union from rasa.engine.graph import GraphComponent, ExecutionContext from rasa.engine.recipes.default\_recipe import DefaultV1Recipe from rasa.engine.storage.resource import Resource from rasa.engine.storage.storage import ModelStorage from rasa.shared.nlu.training\_data.message import Message from rasa.shared.nlu.training\_data.training\_data import TrainingData import ktrain

@DefaultV1Recipe.register( [DefaultV1Recipe.ComponentType.INTENT\_CLASSIFIER], is\_trainable=True )

class KtrainIntentClassifier(GraphComponent):

```
def __init__ (self, component_config: Dict[str, Any]):
    super(). __init__ (component_config)
    # load your pre-trained model here
    self.model = ktrain.load_predictor("distil_bert/distilbert_model_40epochs")

def train(self, training_data, cfg, **kwargs) -> Dict[str, Any]:
    # this component doesn't need to be trained as it's pre-trained
    return {"model_file": "distil_bert/distilbert_model_40epochs"}

def process(self, message, **kwargs):
    # use the pre-trained model to predict the intent of the message
    intent = self.model.predict(message.text)[0]
    message.set("intent", {"name": intent, "confidence": 1.0}, add_to_output=True)

@classmethod
def load(cls, model_dir=None, model_metadata=None, cached_component=None, **kwargs):
    model_file = model_metadata.get("model_file")
    component = cls(model_metadata)
    component.model = ktrain.load_predictor(model_file)
    return component

def persist(self, file_name: Text, model_dir: Text) -> Dict[Text, Any]:
    pass

```

---

<div class="post-metadata">

**Author:** ![gayathrik](https://avatars.discourse-cdn.com/v4/letter/g/958977/32.png) [@gayathrik](https://forum.rasa.com/u/gayathrik)\
**Post date:** [January 19, 2023, 8:05am UTC](https://forum.rasa.com/t/rasa-server-2023-01-19-0756-error-rasa-core-agent-could-not-load-model-due-to-error-initializing-graph-component-for-node-run-languagemodelfeaturizer1/56910/2 "2023-01-19T08:05:24Z")

</div>

can anybody, please help me with correcting the code, using RASA 3 version

---

<div class="post-metadata">

**Author:** ![parisapouya](https://avatars.discourse-cdn.com/v4/letter/p/73ab20/32.png) [@parisapouya](https://forum.rasa.com/u/parisapouya)\
**Post date:** [March 7, 2023, 12:01am UTC](https://forum.rasa.com/t/rasa-server-2023-01-19-0756-error-rasa-core-agent-could-not-load-model-due-to-error-initializing-graph-component-for-node-run-languagemodelfeaturizer1/56910/3 "2023-03-07T00:01:31Z")

</div>

Hi! Did you find a solution for this? And, what LanguageModelFeaturizer are you using?