Hi @nik202 hope you are doing well. I’ve been successfully running rasa open source using docker-compose on cloud server. The thing is I needed to update credentials.yml file to enable web socket channel so I built new image with changes made but duckling issue is happening again. I checked all the configuration and it looked same as one that I used before, but still failing. Only the image I built last time is working without any issue. I tried removing all the images and volumes on the cloud server but error still exists. What’s wrong with this case? I can only think about the duckling endpoint in config.yml url: http://duckling:8000
Is there possibility to cache this value with url: http://localhost:8000
when new image is being built for some reason? I also tried deleting all the image and built new image with new tag with right configuration but it’s not working. Any thoughts?
@jpark2111 Hello, and thanks for the concern. Yes, by god grace I’m doing fine, I wish and hope you also doing fine?
Can you please share a few pieces of information?
- Rasa open source docker and docker-compose files and which version of rasa?
- Errors screenshot, please
- config.yml
Weill solves this issue together.
@nik202 Thanks for quick response on weekends! I’ve been doing great, glad to hear you are doing fine as well
sharing files you requested
- dockerfile
FROM rasa/rasa:2.8.2
WORKDIR /app
COPY . /app
USER root
RUN pip install --upgrade pip
RUN pip install --no-cache-dir -r requirements.txt
RUN python -m spacy download en_core_web_md
RUN python -m spacy link en_core_web_md en
COPY ./data /app/data
COPY config.yml /app
COPY endpoints.yml /app
RUN rasa train
VOLUME /app
VOLUME /app/data
VOLUME /app/models
CMD ["run", "-m", "/app/models", "--enable-api","--cors","*","--debug" ,"--endpoints", "endpoints.yml", "--log-file", "out.log", "--debug"]
EXPOSE 5005
- docker-compose.yml
version: "3.4"
services:
rasa:
restart: always
image: "image:tag"
user: root
ports:
- 5005:5005
volumes:
- ./:/app
depends_on:
- duckling
- app
app:
restart: always
image: "image:tag"
ports:
- 5055:5055
duckling:
restart: always
image: "rasa/duckling:latest"
ports:
- 8000:8000
command: ["duckling-example-exe", "-p", "8000", "--no-access-log", "--no-error-log"]
- Errors Screenshot
- Config.yml
version: "2.0"
language: en
pipeline:
- name: WhitespaceTokenizer
- name: RegexFeaturizer
- name: LexicalSyntacticFeaturizer
- name: CountVectorsFeaturizer
- name: CountVectorsFeaturizer
analyzer: "char_wb"
min_ngram: 1
max_ngram: 4
- name: DIETClassifier
epochs: 100
- name: FallbackClassifier
threshold: 0.7
- name: DucklingEntityExtractor
url: http://duckling:8000
dimensions:
- amount-of-money
- time
- number
- name: SpacyNLP
model: "en_core_web_md"
case_sensitive: false
- name: "SpacyEntityExtractor"
# Note: It is not possible to use the SpacyTokenizer + SpacyFeaturizer in
# combination with the WhitespaceTokenizer, and as a result the
# PERSON extraction by Spacy is not very robust.
# Because of this, the nlu training data is annotated as well, and the
# DIETClassifier will also extract PERSON entities .
dimensions: ["PERSON"]
- name: EntitySynonymMapper
policies:
- name: AugmentedMemoizationPolicy
- name: TEDPolicy
epochs: 40
- name: RulePolicy
core_fallback_threshold: 0.4
core_fallback_action_name: "action_default_fallback"
enable_fallback_prediction: True
@jpark2111 please share docker ps
OR
@jpark2111
Mentioned this in config.yml
(replace these two fields)
- name: "DucklingHTTPExtractor"
url: "http://duckling:8000"
please run it again and share the result or error?
@jpark2111 please make the changes and do let me know, what I have mentioned in my previous post. Failed to connect duckling server while using docker-compose - #4 by nik202
@nik202 So change it with doubleQuotes? will do, it will need time to build image and docker-compose down and up. Should I remove all the images and volumes as well?
@jpark2111 yes, use double quotes please, and yes it will be best if you delete the image and volumes. I hope you are using docker desktop for the same or docker extension in VS code
@nik202 Still getting same error. wondering why it worked before, and now it doesn’t with same configuration…
@nik202 duckling is the only thing running on port 8000, but is there a way that I can check to make sure duckling is the only one running on port 8000?
@jpark2111 I guess that is fine, can you mention EXPOSE 8000 in docker-compose.yml under ducking?
expose:
- "8000"
and run the image
@jpark2111 is your code is running on the local machine without any issue? did you cross-check?
@jpark2111 quickly share the endpoints.yml
expose:
- "8000"
still getting error.
locally it’s always working fine. @nik202
endpoints.yml
version: "2.0"
action_endpoint:
url: "http://app:5055/webhook"
There’s only action server endpoint configured.