-
Notifications
You must be signed in to change notification settings - Fork 0
Deliverable 3: MVP

- Added the gunicorn and nginx to Programs.
- Added "dall-e-2" model to API node along gpt-3.5-turbo.
- Specified the PostgreSQL version.
- Specified operating version system (was "Ubuntu Jammy", now is "Ubuntu 22.04.1").
- Switched from an AWS EC2 instance to an Oracle Cloud instance.
- Updated libraries:
- Removed pandas and pillow.
- Specified matplotlib version.
- Added dotenv, openai, and django-bootstrap.
- Changed "ChatGPT 3.5" to "gpt-3.5-turbo".
- Downgraded to Python 3.10.12 from Python 3.11.1.

We use an MVC architecture where each component is, for the most part, directly mapped to a Django "app".
- Authentication: Handles the sign-up, login, logout, and session.
- Recipes: Generates and displays recipes and their corresponding images.
- Kitchen: Creates ingredients and reminds user to update it.
- Statistics: Generates and shows a report.
- User: Handles editing and deleting accounts.
No changes.

- Fixed the relationship between
Ingredient.user_idandUser.id(wasUser.username). - Removed the fields
first_name,last_name,is_staff,is_active,is_superuser,date_joinedfrom the User table as they were not that important to represent our system, but they still exist as fields in Django's default User model. - Removed the fields
unit_countandmeasurement_unitfrom the Ingredient table.
| # | Column | Data type | Description |
|---|---|---|---|
| 1 | id | int | Unique id. |
| 2 | varchar | Used for log-in. | |
| 3 | password | hash | Log-in credentials. Not stored as plain-text. |
| 4 | last_login | date | Last time the user interacted with the system. |
| # | Column | Data type | Description |
|---|---|---|---|
| 1 | user_id | int | Id of the user. |
| 2 | last_modified_at | timestamp | Unix time when the user last modified any ingredients. |
| # | Column | Data type | Description |
|---|---|---|---|
| 1 | id | int | Unique id. |
| 2 | user_id | int | Id of the user for whom the recipe was generated. |
| 3 | title | varchar | Title generated by the LLM. |
| 4 | description | varchar | Steps of the recipe generated by the LLM. |
| 5 | prompt | varchar | Prompt sent to the LLM to generate the recipe |
| 6 | generated_at | timestamp | Unix time when the recipe was generated. |
| 7 | is_favorite | boolean | Whether the user marked the recipe as favorite. |
| # | Column | Data type | Description |
|---|---|---|---|
| 1 | id | int | Unique id. |
| 2 | user_id | int | Id of the user that created the ingredient. |
| 3 | name | varchar | Name of the ingredient (user inputted). |
| # | Column | Data type | Description |
|---|---|---|---|
| 1 | recipe_id | int | Id of a generated recipe. |
| 2 | ingredient_name | varchar | Name of the ingredient at the time the recipe was generated. |

A complementary .ase file can be downloaded from
here.

Glory and Montserrat are both licensed under the Open Font License.

Generated with Microsoft Image Creator.

We want to emphasize that our app uses the ingredients you already have so that they don't go to waste.



-
Generated with the
pipreqspackage (which only includes packages that have been imported in the project). -
README.mdwith installation instructions.
https://www.youtube.com/watch?v=dQw4w9WgXcQ
Sprint backlog: https://github.com/users/alejoriosm04/projects/1/views/3?filterQuery=iteration%3A%22Iteration+3%22
FridgeToFeast is a web application that helps users create personalized recipes using the ingredients available in their fridge and pantry through AI.
- Deliverable 1: Project definition and Software Requirements Specification
- Deliverable 2: Solution design and Funcionality preview
- Deliverable 3: MVP
- Deliverable 4: MVP + Documentation
- Weekly meetings
- Weekly 1: 2024‐01‐30
- Weekly 2: 2024‐02‐07
- Weekly 3: 2024‐02‐12
- Weekly 4: 2024‐02‐22
- Weekly 5: 2024‐02‐27
- Weekly 6: 2024‐03‐05
- Weekly 7: 2024‐03‐11
- Weekly 8: 2024‐03‐19
- Weekly 9: 2024‐04‐04
- Weekly 10: 2024‐04‐11
- Weekly 11: 2024‐04‐16
- Weekly 12: 2024‐04‐23
- Weekly 13: 2024‐05‐02
- Weekly 14: 2024‐05‐07
- Weekly 15: 2024‐05‐14
- Retrospective