summarization¶

• Main author - HE-Arc
• Code
• Deployment configuration
• Staging
• Production

Description¶

This service uses langchain and LLMs to create summaries of documents. The technique used for the summarization is the Map/reduce technique showed by langchain. Each document will be summarized individually (Mapping phase) and then all the summaries will be combined into a single summary (Reduce phase).

The LLMs used by this service are hosted by Ollama.

Environment variables¶

Check the Core concepts > Service > Environment variables documentation for more details.

Run the tests with Python¶

Check the Core concepts > Service > Run the tests with Python documentation for more details.

Start the service locally¶

Check the Core concepts > Service > Start the service locally documentation for more details.

Deploy ollama locally¶

add the following container to the stateful.yml deployment file

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19

- name: ollama
image: ollama/ollama
ports:
- name: http
    containerPort: 11434
command: ["/bin/bash", "-c"]
args:
- |
    ollama serve &
    sleep 10
    ollama pull mistral:instruct
    sleep infinity
resources:
    requests:
    tencent.com/vcuda-core: 20
    tencent.com/vcuda-memory: 8
    limits:
    tencent.com/vcuda-core: 20
    tencent.com/vcuda-memory: 8