Skip to content

Commit

Permalink
🦙
Browse files Browse the repository at this point in the history
  • Loading branch information
@hardbyte
hardbyte committed Feb 6, 2025
1 parent 9ecb9a3 commit c792abd
Show file tree
Hide file tree
Showing 2 changed files with 49 additions and 97 deletions.
133 changes: 42 additions & 91 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -13,61 +13,6 @@ Install with `uv`, `pipx`, `pip` etc:
uv tool install qabot
```

## Security Risks

This program gives an LLM access to your local and network accessible files and allows it to execute arbitrary SQL
queries in a DuckDB database, see [Security](Security.md) for more information.

## Command Line Usage

```bash
$ EXPORT OPENAI_API_KEY=sk-...
$ qabot -f data/titanic.csv -q "How many passengers were there by ticket class and gender?"
🤖 There were a total of 891 passengers, categorized by ticket class and gender.


The distribution of passengers by ticket class and gender is as follows:

| Ticket Class | Gender | Passenger Count |
|--------------|--------|------------------|
| 1 | female | 94 |
| 1 | male | 122 |
| 2 | female | 76 |
| 2 | male | 108 |
| 3 | female | 144 |
| 3 | male | 347 |

This was computed by grouping the data in the 'titanic' table by 'Pclass' and 'Sex' and counting the number of passengers in each
category.

SELECT Pclass, Sex, COUNT(*) AS PassengerCount FROM titanic GROUP BY Pclass, Sex ORDER BY Pclass, Sex;
```

## LLM Providers

qabot works with any OpenAI compatible api including deepseek. Simple set the base URL:
```
export OPENAI_BASE_URL=https://api.deepseek.com
```

## Python Usage

```python
from qabot import ask_wikidata, ask_file, ask_database

print(ask_wikidata("How many hospitals are there in New Zealand?"))
print(ask_file("How many men were aboard the titanic?", 'data/titanic.csv'))
print(ask_database("How many product images are there?", 'postgresql://user:password@localhost:5432/dbname'))
```

Output:
```text
There are 54 hospitals in New Zealand.
There were 577 male passengers on the Titanic.
There are 6,225 product images.
```


## Features

Works on local CSV, sqlite and Excel files:
Expand Down Expand Up @@ -156,36 +101,59 @@ Install the `qabot` command line tool using uv/pip/pipx:
$ uv tool install qabot
```

Then run the `qabot` command with either local files (`-f my-file.csv`) or `-w` to query wikidata.
Then run the `qabot` command with optional files (`-f my-file.csv`) and an initial query `-q "How many..."`.

See all options with `qabot --help`

## Examples
## Security Risks

### Local CSV file/s
This program gives an LLM access to your local and network accessible files and allows it to execute arbitrary SQL
queries in a DuckDB database, see [Security](Security.md) for more information.

```bash
$ qabot -q "how many passengers survived by gender?" -f data/titanic.csv
🦆 Loading data from files...
Loading data/titanic.csv into table titanic...

Query: how many passengers survived by gender?
Result:
There were 233 female passengers and 109 male passengers who survived.
## LLM Providers

qabot works with any OpenAI compatible api including Ollama and deepseek. Simple set the base URL:
```
export OPENAI_BASE_URL=https://api.deepseek.com
```

🚀 any further questions? [y/n] (y): y
Or Ollama:
```
OPENAI_BASE_URL=http://localhost:11434/v1/
QABOT_MODEL_NAME=qwen2.5-coder:7b
QABOT_PLANNING_MODEL_NAME=deepseek-r1:14b
```

🚀 Query: what was the largest family who did not survive?
Query: what was the largest family who did not survive?
Result:
The largest family who did not survive was the Sage family, with 8 members.
## Python API

🚀 any further questions? [y/n] (y): Show the survival rate by gender, and ticket class shown as an ASCII graph
```python
from qabot import ask_wikidata, ask_file, ask_database

Here’s the survival count represented as a horizontal bar graph grouped by ticket class and gender:
print(ask_wikidata("How many hospitals are there in New Zealand?"))
print(ask_file("How many men were aboard the titanic?", 'data/titanic.csv'))
print(ask_database("How many product images are there?", 'postgresql://user:password@localhost:5432/dbname'))
```

Output:
```text
There are 54 hospitals in New Zealand.
There were 577 male passengers on the Titanic.
There are 6,225 product images.
```


## Examples

### Local CSV file/s

```bash
$ qabot -q "Show the survival rate by gender, and ticket class shown as an ASCII graph" -f data/titanic.csv
🦆 Loading data from files...
Loading data/titanic.csv into table titanic...

Here’s the survival count represented as a horizontal bar graph grouped by ticket class and gender:

Class 1:
Females | ██████████████████████████████████████████ (91)
Males | ██████████████ (45)
Expand All @@ -200,7 +168,6 @@ Males | ██████████████ (47)


This representation allows us to observe that in all classes, a greater number of female passengers survived compared to male passengers, and also highlights the number of survivors is notably higher in the first class compared to the other classes.

```


Expand All @@ -217,10 +184,6 @@ $ qabot -w -q "How many Hospitals are there located in Beijing"
Use the `-v` flag to see the intermediate steps and database queries.
Sometimes it takes a long route to get to the answer, but it's often interesting to see how it gets there.

```
qabot -f data/titanic.csv -q "how many passengers survived by gender?" -v
```

## Data accessed via http/s3

Use the `-f <url>` flag to load data from a url, e.g. a csv file on s3:
Expand All @@ -236,19 +199,7 @@ Result:
## Docker Usage
You can build and run the Docker image for `qabot` using the following instructions:
### Building the Docker Image
To build the Docker image, run the following command in the root directory of the repository:
```bash
docker build -t qabot .
```
### Running the Docker Image
To run the Docker image, use the following command:
You can run `qabot` via Docker:
```bash
docker run --rm \
Expand All @@ -260,7 +211,7 @@ docker run --rm \
Replace the mount path to your actual data along with replacing `your_openai_api_key`.
## Ideas
- G-Sheets via https://github.com/evidence-dev/duckdb_gsheets
- Streaming mode to output results as they come in
- token limits and better reporting of costs
- Supervisor agent - assess whether a query is "safe" to run, could ask for user confirmation to run anything that gets flagged.
Expand Down
13 changes: 7 additions & 6 deletions qabot/cli.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -201,13 +201,14 @@ def terminate_session(message: str):
print(format_rocket("Question:"))
print(format_user(query))

print(format_robot(result["summary"]))
print()
if "detail" in result:
print(f"[{ROBOT_COLOR}]\n{result['detail']}\n")
if result:
print(format_robot(result["summary"]))
print()
if "detail" in result:
print(f"[{ROBOT_COLOR}]\n{result['detail']}\n")

if "query" in result:
print(format_query(result["query"]))
if "query" in result:
print(format_query(result["query"]))

print()
query = Prompt.ask(FOLLOW_UP_PROMPT)
Expand Down

0 comments on commit c792abd

Please sign in to comment.