You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+20-20
Original file line number
Diff line number
Diff line change
@@ -1,9 +1,9 @@
1
1
2
2
# 🍊 tangerine (backend) <!-- omit from toc -->
3
3
4
-
tangerine is a slim and light-weight RAG (Retieval Augmented Generated) system used to create and manage chat bot agents.
4
+
tangerine is a slim and light-weight RAG (Retieval Augmented Generated) system used to create and manage chat bot assistants.
5
5
6
-
Each agent is intended to answer questions related to a set of documents known as a knowledge base (KB).
6
+
Each assistant is intended to answer questions related to a set of documents known as a knowledge base (KB).
7
7
8
8

9
9
@@ -61,10 +61,10 @@ It was born out of a hack-a-thon and is still a work in progress. You will find
61
61
62
62
#### Retrieval Augmented Generation (RAG)
63
63
64
-
-**1:** A user presents a question to an agent in the chat interface
64
+
-**1:** A user presents a question to an assistant in the chat interface
65
65
-**2:** Embeddings are created for the query using the embedding model
66
66
-**3:** A similarity search and a max marginal relevance search are performed against the vector DB to find the top N most relevant document chunks
67
-
- The document set searched is scoped only to that specific agent
67
+
- The document set searched is scoped only to that specific assistant
68
68
-**4:** The LLM is prompted to answer the question using only the context found within the relevant document chunks
69
69
-**5:** The LLM response is streamed by the backend service to the user. Metadata containing the document chunks are also returned to be used as citations.
70
70
@@ -110,7 +110,7 @@ Our documentation set has initially focused on pages that have been compiled usi
110
110
111
111
The **tangerine-backend** service manages:
112
112
113
-
- Create/update/delete of chat bot "agents" via REST API.
113
+
- Create/update/delete of chat bot "assistants" via REST API.
114
114
- Document ingestion
115
115
- Upload via the API, or sync via an s3 bucket
116
116
- Text cleanup/conversion
@@ -176,7 +176,7 @@ The docker compose file offers an easy way to spin up all components. [ollama](h
176
176
6. Access the API on port `8000`
177
177
178
178
```sh
179
-
curl -XGET 127.0.0.1:8000/api/agents
179
+
curl -XGET 127.0.0.1:8000/api/assistants
180
180
{
181
181
"data": []
182
182
}
@@ -293,7 +293,7 @@ to use this to test different embedding models that are not supported by ollama,
293
293
1. Access the API on port `8000`
294
294
295
295
```sh
296
-
curl -XGET 127.0.0.1:8000/api/agents
296
+
curl -XGET 127.0.0.1:8000/api/assistants
297
297
{
298
298
"data": []
299
299
}
@@ -376,7 +376,7 @@ Comment out `ollama` from the compose file, or stop the ollama container. Invoke
376
376
377
377
## Synchronizing Documents from S3
378
378
379
-
You can configure a set of agents and continually sync their knowledge base via documents stored in an S3 bucket.
379
+
You can configure a set of assistants and continually sync their knowledge base via documents stored in an S3 bucket.
380
380
381
381
To do so you'll need to do the following:
382
382
@@ -400,7 +400,7 @@ To do so you'll need to do the following:
400
400
echo 'BUCKET=mybucket' >> .env
401
401
```
402
402
403
-
1. Create an `s3.yaml` file that describes your agents and the documents they should ingest. See [s3-example.yaml](s3-example.yaml) for an example.
403
+
1. Create an `s3.yaml` file that describes your assistants and the documents they should ingest. See [s3-example.yaml](s3-example.yaml) for an example.
404
404
405
405
If using docker compose, copy this config into your container:
406
406
@@ -422,7 +422,7 @@ To do so you'll need to do the following:
422
422
flask s3sync
423
423
```
424
424
425
-
The sync creates agents and ingests the configured documents for each agent. After initial creation, when the task is run it checks the S3 bucket for updates and will only re-ingest files into the vector DB when it detects file changes.
425
+
The sync creates assistants and ingests the configured documents for each assistant. After initial creation, when the task is run it checks the S3 bucket for updates and will only re-ingest files into the vector DB when it detects file changes.
426
426
427
427
The OpenShift templates contain a CronJob configuration that is used to run this document sync repeatedly.
428
428
@@ -445,19 +445,19 @@ This repository provides [OpenShift templates](openshift/) for all infrastructur
445
445
446
446
## Run Tangerine Frontend Locally
447
447
448
-
The API can be used to create/manage/update agents, upload documents, and to chat with each agent. However, the frontend provides a simpler interface to manage the service with. To run the UI in a development environment, see [tangerine-frontend](https://github.com/RedHatInsights/tangerine-frontend)
448
+
The API can be used to create/manage/update assistants, upload documents, and to chat with each assistant. However, the frontend provides a simpler interface to manage the service with. To run the UI in a development environment, see [tangerine-frontend](https://github.com/RedHatInsights/tangerine-frontend)
0 commit comments