Cyril
/
ai-lab-transformers-playground
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Update README and configuration files; add Question Answering to CLI and API, and adjust default models

This commit is contained in:
Cyril 2025-10-12 22:08:57 +02:00
parent 8fb16726d6
commit f081ef0db0
5 changed files with 167 additions and 335 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
.gitignore vendored
View File

@ -22,6 +22,7 @@ var/
*.egg-info/ *.egg-info/
.installed.cfg .installed.cfg
*.egg *.egg
.github/
# Virtual environments # Virtual environments
venv/ venv/

472
README.md
View File

@ -1,16 +1,54 @@
# 🧠 AI Lab – Transformers CLI Playground # 🧠 AI Lab – Transformers CLI Playground
> A **pedagogical and technical project** designed for AI practitioners and students to experiment with Hugging Face Transformers through an **interactive Command‑Line Interface (CLI)**. > A **pedagogical and technical project** designed for AI practitioners and students to explore **Hugging Face Transformers** through an **interactive Command-Line Interface (CLI)** or a **REST API**.
> This playground provides ready‑to‑use NLP pipelines (Sentiment Analysis, Named Entity Recognition, Text Generation, Fill‑Mask, Moderation, etc.) in a modular, extensible, and educational codebase. > This playground provides ready-to-use NLP pipelines — including **Sentiment Analysis**, **Named Entity Recognition**, **Text Generation**, **Fill-Mask**, **Question Answering (QA)**, **Moderation**, and more — in a modular, extensible, and educational codebase.
---
<p align="center">
<img src="https://img.shields.io/badge/Python-3.13-blue.svg" alt="Python"/>
<img src="https://img.shields.io/badge/Built_with-Poetry-purple.svg" alt="Poetry"/>
<img src="https://img.shields.io/badge/🤗-Transformers-orange.svg" alt="Transformers"/>
<img src="https://img.shields.io/badge/License-MIT-green.svg" alt="License"/>
</p>
---
## 📑 Table of Contents
- [📚 Overview](#-overview)
- [🗂️ Project Structure](#️-project-structure)
- [⚙️ Installation](#️-installation)
- [🧾 Option 1 – Poetry (Recommended)](#-option-1--poetry-recommended)
- [📦 Option 2 – Pip + Requirements](#-option-2--pip--requirements)
- [▶️ Usage](#️-usage)
- [🖥️ CLI Mode](#️-cli-mode)
- [🌐 API Mode](#-api-mode)
- [📡 API Endpoints](#-api-endpoints)
- [🖥️ CLI Examples](#️-cli-examples)
- [🧠 Architecture Overview](#-architecture-overview)
- [⚙️ Configuration](#️-configuration)
- [🧩 Extending the Playground](#-extending-the-playground)
- [🧰 Troubleshooting](#-troubleshooting)
- [🧭 Development Guidelines](#-development-guidelines)
- [🧱 Roadmap](#-roadmap)
- [📜 License](#-license)
--- ---
## 📚 Overview ## 📚 Overview
The **AI Lab – Transformers CLI Playground** allows you to explore multiple natural language processing tasks directly from the terminal. The **AI Lab – Transformers CLI Playground** enables users to explore **multiple NLP tasks directly from the terminal or via HTTP APIs**.
Each task (e.g., sentiment, NER, text generation) is implemented as a **Command Module**, which interacts with a **Pipeline Module** built on top of the `transformers` library. Each task (sentiment, NER, text generation, etc.) is implemented as a **Command Module** that communicates with a **Pipeline Module** powered by Hugging Face’s `transformers` library.
The lab is intentionally structured to demonstrate **clean software design for ML codebases** — with strict separation between configuration, pipelines, CLI logic, and display formatting. The project demonstrates **clean ML code architecture** with strict separation between:
- Configuration
- Pipelines
- CLI logic
- Display formatting
It’s a great educational resource for learning **how to structure ML applications** professionally.
--- ---
@ -18,61 +56,54 @@ The lab is intentionally structured to demonstrate **clean software design for M
```text ```text
src/ src/
├── __init__.py
├── main.py # CLI entry point ├── main.py # CLI entry point
│ │
├── cli/ ├── cli/
│ ├── __init__.py │ ├── base.py # CLICommand base class & interactive shell
│ ├── base.py # CLICommand base class & interactive shell handler │ └── display.py # Console formatting utilities (colors, tables, results)
│ └── display.py # Console formatting utilities (tables, colors, results)
│ │
├── commands/ # User-facing commands wrapping pipeline logic ├── commands/ # User-facing commands wrapping pipeline logic
│ ├── __init__.py
│ ├── sentiment.py # Sentiment analysis command │ ├── sentiment.py # Sentiment analysis command
│ ├── fillmask.py # Masked token prediction command │ ├── fillmask.py # Masked token prediction
│ ├── textgen.py # Text generation command │ ├── textgen.py # Text generation
│ ├── ner.py # Named Entity Recognition command │ ├── ner.py # Named Entity Recognition
│ └── moderation.py # Toxicity / content moderation command │ ├── qa.py # Question Answering (extractive)
│ └── moderation.py # Content moderation / toxicity detection
│ │
├── pipelines/ # Machine learning logic (Hugging Face Transformers) ├── pipelines/ # ML logic based on Hugging Face pipelines
│ ├── __init__.py
│ ├── template.py # Blueprint for creating new pipelines │ ├── template.py # Blueprint for creating new pipelines
│ ├── sentiment.py │ ├── sentiment.py
│ ├── fillmask.py │ ├── fillmask.py
│ ├── textgen.py │ ├── textgen.py
│ ├── ner.py │ ├── ner.py
│ ├── qa.py
│ └── moderation.py │ └── moderation.py
│ │
├── api/ ├── api/
│ ├── __init__.py │ ├── app.py # FastAPI app and endpoints
│ ├── app.py # FastAPI application with all endpoints │ ├── models.py # Pydantic schemas
│ ├── models.py # Pydantic request/response models │ └── config.py # API configuration
│ └── config.py # API-specific configuration
│ │
└── config/ └── config/
├── __init__.py └── settings.py # Global configuration (models, params)
└── settings.py # Global configuration (default models, parameters)
``` ```
--- ---
## ⚙️ Installation ## ⚙️ Installation
### 🧾 Option 1 – Using Poetry (Recommended) ### 🧾 Option 1 – Poetry (Recommended)
> Poetry is used as the main dependency manager. > Poetry is the main dependency manager for this project.
```bash ```bash
# 1. Create and activate a new virtual environment
poetry shell poetry shell
# 2. Install dependencies
poetry install poetry install
``` ```
This will automatically install all dependencies declared in `pyproject.toml`, including **transformers**, **torch**, and **FastAPI** for the API mode. This installs all dependencies defined in `pyproject.toml` (including `transformers`, `torch`, and `fastapi`).
To run the application inside the Poetry environment: Run the app:
```bash ```bash
# CLI mode # CLI mode
@ -84,21 +115,15 @@ poetry run python src/main.py --mode api
--- ---
### 📦 Option 2 – Using pip and requirements.txt ### 📦 Option 2 – Pip + requirements.txt
If you prefer using `requirements.txt` manually: If you prefer manual dependency management:
```bash ```bash
# 1. Create a virtual environment
python -m venv .venv python -m venv .venv
source .venv/bin/activate # Linux/macOS
.venv\Scripts\Activate.ps1 # Windows
# 2. Activate it
# Linux/macOS
source .venv/bin/activate
# Windows PowerShell
.venv\Scripts\Activate.ps1
# 3. Install dependencies
pip install -r requirements.txt pip install -r requirements.txt
``` ```
@ -106,19 +131,15 @@ pip install -r requirements.txt
## ▶️ Usage ## ▶️ Usage
The application supports two modes: **CLI** (interactive) and **API** (REST server).
### 🖥️ CLI Mode ### 🖥️ CLI Mode
Launch the interactive CLI with: Run the interactive CLI:
```bash ```bash
python -m src.main --mode cli python -m src.main --mode cli
# or, if using Poetry
poetry run python src/main.py --mode cli
``` ```
You'll see an interactive menu listing the available commands: Interactive menu:
``` ```
Welcome to AI Lab - Transformers CLI Playground Welcome to AI Lab - Transformers CLI Playground
@ -127,151 +148,89 @@ Available commands:
• fillmask – Predict masked words in a sentence • fillmask – Predict masked words in a sentence
• textgen – Generate text from a prompt • textgen – Generate text from a prompt
• ner – Extract named entities from text • ner – Extract named entities from text
• qa – Answer questions from a context
• moderation – Detect toxic or unsafe content • moderation – Detect toxic or unsafe content
``` ```
---
### 🌐 API Mode ### 🌐 API Mode
Launch the FastAPI server with: Run FastAPI server:
```bash ```bash
python -m src.main --mode api python -m src.main --mode api
# or with custom settings # Custom config
python -m src.main --mode api --host 0.0.0.0 --port 8000 --reload python -m src.main --mode api --host 0.0.0.0 --port 8000 --reload
``` ```
The API will be available at: API Docs:
- **Swagger Documentation**: http://localhost:8000/docs - **Swagger** → http://localhost:8000/docs
- **ReDoc Documentation**: http://localhost:8000/redoc - **ReDoc** → http://localhost:8000/redoc
- **OpenAPI Schema**: http://localhost:8000/openapi.json - **OpenAPI** → http://localhost:8000/openapi.json
---
## 📡 API Endpoints ## 📡 API Endpoints
The REST API provides all CLI functionality through HTTP endpoints:
### Core Endpoints ### Core Endpoints
| Method | Endpoint | Description | | Method | Endpoint | Description |
| ------ | --------- | -------------------------------- | | ------ | --------- | ------------------------- |
| `GET` | `/` | Health check and API information | | `GET` | `/` | Health check and API info |
| `GET` | `/health` | Detailed health status | | `GET` | `/health` | Detailed health status |
### Individual Processing ### Individual Processing
| Method | Endpoint | Description | Input | | Method | Endpoint | Description |
| ------ | ------------- | ------------------------ | ------------------------------------------------------------------ | | ------ | ------------- | ---------------------- |
| `POST` | `/sentiment` | Analyze text sentiment | `{"text": "string", "model": "optional"}` | | `POST` | `/sentiment` | Analyze text sentiment |
| `POST` | `/fillmask` | Fill masked words | `{"text": "Hello [MASK]", "model": "optional"}` | | `POST` | `/fillmask` | Predict masked words |
| `POST` | `/textgen` | Generate text | `{"text": "prompt", "model": "optional"}` | | `POST` | `/textgen` | Generate text |
| `POST` | `/ner` | Named entity recognition | `{"text": "string", "model": "optional"}` | | `POST` | `/ner` | Extract named entities |
| `POST` | `/qa` | Question answering | `{"question": "string", "context": "string", "model": "optional"}` | | `POST` | `/qa` | Question answering |
| `POST` | `/moderation` | Content moderation | `{"text": "string", "model": "optional"}` | | `POST` | `/moderation` | Content moderation |
### Batch Processing ### Batch Processing
| Method | Endpoint | Description | Input | | Method | Endpoint | Description |
| ------ | ------------------- | ------------------------------------ | ---------------------------------------------------- | | ------ | ------------------- | -------------------------- |
| `POST` | `/sentiment/batch` | Process multiple texts | `{"texts": ["text1", "text2"], "model": "optional"}` | | `POST` | `/sentiment/batch` | Process multiple texts |
| `POST` | `/fillmask/batch` | Fill multiple masked texts | `{"texts": ["text1 [MASK]"], "model": "optional"}` | | `POST` | `/fillmask/batch` | Fill multiple masked texts |
| `POST` | `/textgen/batch` | Generate from multiple prompts | `{"texts": ["prompt1"], "model": "optional"}` | | `POST` | `/textgen/batch` | Generate from prompts |
| `POST` | `/ner/batch` | Extract entities from multiple texts | `{"texts": ["text1"], "model": "optional"}` | | `POST` | `/ner/batch` | Extract entities in batch |
| `POST` | `/moderation/batch` | Moderate multiple texts | `{"texts": ["text1"], "model": "optional"}` | | `POST` | `/qa/batch` | Answer questions in batch |
| `POST` | `/moderation/batch` | Moderate multiple texts |
### Example API Usage
#### 🔹 Sentiment Analysis
```bash
curl -X POST "http://localhost:8000/sentiment" \
-H "Content-Type: application/json" \
-d '{"text": "I absolutely love this project!"}'
```
Response:
```json
{
"success": true,
"label": "POSITIVE",
"score": 0.998,
"model_used": "distilbert-base-uncased-finetuned-sst-2-english"
}
```
#### 🔹 Named Entity Recognition
```bash
curl -X POST "http://localhost:8000/ner" \
-H "Content-Type: application/json" \
-d '{"text": "Elon Musk founded SpaceX in California."}'
```
Response:
```json
{
"success": true,
"entities": [
{ "word": "Elon Musk", "label": "PERSON", "score": 0.999 },
{ "word": "SpaceX", "label": "ORG", "score": 0.998 },
{ "word": "California", "label": "LOC", "score": 0.995 }
],
"model_used": "dslim/bert-base-NER"
}
```
#### 🔹 Batch Processing
```bash
curl -X POST "http://localhost:8000/sentiment/batch" \
-H "Content-Type: application/json" \
-d '{"texts": ["Great product!", "Terrible experience", "It was okay"]}'
```
Response:
```json
{
"success": true,
"results": [
{ "label": "POSITIVE", "score": 0.998 },
{ "label": "NEGATIVE", "score": 0.995 },
{ "label": "NEUTRAL", "score": 0.876 }
],
"model_used": "distilbert-base-uncased-finetuned-sst-2-english"
}
```
--- ---
## 🖥️ CLI Examples ## 🖥️ CLI Examples
#### 🔹 Sentiment Analysis ### 🔹 Sentiment Analysis
```text ```text
💬 Enter text: I absolutely love this project! 💬 Enter text: I absolutely love this project!
→ Sentiment: POSITIVE (score: 0.998) → Sentiment: POSITIVE (score: 0.998)
``` ```
#### 🔹 Fill‑Mask ### 🔹 Fill-Mask
```text ```text
💬 Enter text: The capital of France is [MASK]. 💬 Enter text: The capital of France is [MASK].
→ Predictions: → Predictions:
1) Paris score: 0.87 1) Paris score: 0.87
2) Lyon score: 0.04 2) Lyon score: 0.04
3) London score: 0.02
``` ```
#### 🔹 Text Generation ### 🔹 Text Generation
```text ```text
💬 Prompt: Once upon a time 💬 Prompt: Once upon a time
→ Output: Once upon a time there was a young AI learning to code... → Output: Once upon a time there was a young AI learning to code...
``` ```
#### 🔹 NER (Named Entity Recognition) ### 🔹 NER
```text ```text
💬 Enter text: Elon Musk founded SpaceX in California. 💬 Enter text: Elon Musk founded SpaceX in California.
@ -281,7 +240,15 @@ Response:
- California (LOC) - California (LOC)
``` ```
#### 🔹 Moderation ### 🔹 QA (Question Answering)
```text
💬 Enter question: What is the capital of France?
💬 Enter context: France is a country in Europe. Its capital is Paris.
→ Answer: The capital of France is Paris.
```
### 🔹 Moderation
```text ```text
💬 Enter text: I hate everything! 💬 Enter text: I hate everything!
@ -292,90 +259,34 @@ Response:
## 🧠 Architecture Overview ## 🧠 Architecture Overview
The application supports dual-mode architecture: **CLI** (interactive) and **API** (REST server), both sharing the same pipeline layer: Both CLI and API share the **same pipeline layer**, ensuring code reusability and consistency.
### CLI Architecture ### CLI Architecture
```text ```text
┌──────────────────────┐ InteractiveCLI → Command Layer → Pipeline Layer → Display Layer
│ InteractiveCLI │
│ (src/cli/base.py) │
└──────────┬───────────┘
│
▼
┌─────────────────┐
│ Command Layer │ ← e.g. sentiment.py
│ (user commands) │
└───────┬─────────┘
│
▼
┌─────────────────┐
│ Pipeline Layer │ ← e.g. pipelines/sentiment.py
│ (ML logic) │
└───────┬─────────┘
│
▼
┌─────────────────┐
│ Display Layer │ ← cli/display.py
│ (format output) │
└─────────────────┘
``` ```
### API Architecture ### API Architecture
```text ```text
┌──────────────────────┐ FastAPI App → Pydantic Models → Pipeline Layer → JSON Response
│ FastAPI App │
│ (src/api/app.py) │
└──────────┬───────────┘
│
▼
┌─────────────────┐
│ Pydantic Models │ ← api/models.py
│ (validation) │
└───────┬─────────┘
│
▼
┌─────────────────┐
│ Pipeline Layer │ ← e.g. pipelines/sentiment.py
│ (ML logic) │ (shared with CLI)
└───────┬─────────┘
│
▼
┌─────────────────┐
│ JSON Response │ ← automatic serialization
│ (HTTP output) │
└─────────────────┘
``` ```
### Key Concepts | Layer | Description |
| ------------ | ---------------------------------------------- |
| Layer | Description | | **CLI** | Manages user input/output and navigation. |
| ------------ | -------------------------------------------------------------------------- | | **API** | Exposes endpoints with automatic OpenAPI docs. |
| **CLI** | Manages user input/output, help menus, and navigation between commands. | | **Command** | Encapsulates user-facing operations. |
| **API** | FastAPI application serving HTTP endpoints with automatic documentation. | | **Pipeline** | Wraps Hugging Face’s pipelines. |
| **Command** | Encapsulates a single user-facing operation (e.g., run sentiment). | | **Models** | Validates requests/responses. |
| **Pipeline** | Wraps Hugging Face's `transformers.pipeline()` to perform inference. | | **Display** | Formats console output. |
| **Models** | Pydantic schemas for request/response validation and serialization. |
| **Display** | Handles clean console rendering (colored output, tables, JSON formatting). |
### Key Concepts
| Layer | Description |
| ------------ | -------------------------------------------------------------------------- |
| **CLI** | Manages user input/output, help menus, and navigation between commands. |
| **Command** | Encapsulates a single user-facing operation (e.g., run sentiment). |
| **Pipeline** | Wraps Hugging Face’s `transformers.pipeline()` to perform inference. |
| **Display** | Handles clean console rendering (colored output, tables, JSON formatting). |
| **Config** | Centralizes model names, limits, and global constants. |
--- ---
## ⚙️ Configuration ## ⚙️ Configuration
All configuration is centralized in `src/config/settings.py`. All configuration is centralized in `src/config/settings.py`:
Example:
```python ```python
class Config: class Config:
@ -384,140 +295,75 @@ class Config:
"fillmask": "bert-base-uncased", "fillmask": "bert-base-uncased",
"textgen": "gpt2", "textgen": "gpt2",
"ner": "dslim/bert-base-NER", "ner": "dslim/bert-base-NER",
"moderation":"unitary/toxic-bert" "qa": "distilbert-base-cased-distilled-squad",
"moderation":"unitary/toxic-bert",
} }
MAX_LENGTH = 512 MAX_LENGTH = 512
BATCH_SIZE = 8 BATCH_SIZE = 8
``` ```
You can easily modify model names to experiment with different checkpoints.
--- ---
## 🧩 Extending the Playground ## 🧩 Extending the Playground
To create a new experiment (e.g., keyword extraction): To add a new NLP experiment (e.g., keyword extraction):
### For CLI Support 1. Duplicate `src/pipelines/template.py` → `src/pipelines/keywords.py`
2. Create a command: `src/commands/keywords.py`
3. Register it in `src/main.py`
4. Add Pydantic models and API endpoint
5. Update `Config.DEFAULT_MODELS`
1. **Duplicate** `src/pipelines/template.py` → `src/pipelines/keywords.py` Both CLI and API will automatically share this logic.
Implement the `run()` or `analyze()` logic using a new Hugging Face pipeline.
2. **Create a Command** in `src/commands/keywords.py` to interact with users.
3. **Register the command** inside `src/main.py`:
```python
from src.commands.keywords import KeywordsCommand
cli.register_command(KeywordsCommand())
```
### For API Support
4. **Add Pydantic models** in `src/api/models.py`:
```python
class KeywordsRequest(BaseModel):
text: str
model: Optional[str] = None
class KeywordsResponse(BaseModel):
success: bool
keywords: List[str]
model_used: str
```
5. **Add endpoint** in `src/api/app.py`:
```python
@app.post("/keywords", response_model=KeywordsResponse)
async def extract_keywords(request: KeywordsRequest):
# Implementation using KeywordsAnalyzer pipeline
pass
```
6. **Update configuration** in `Config.DEFAULT_MODELS`.
Both CLI and API will automatically share the same pipeline implementation!
---
## 🧪 Testing
You can use `pytest` for lightweight validation:
```bash
pip install pytest
pytest -q
```
Recommended structure:
```
tests/
├── test_sentiment.py
├── test_textgen.py
└── ...
```
--- ---
## 🧰 Troubleshooting ## 🧰 Troubleshooting
### General Issues | Issue | Solution |
| ------------------------ | ----------------------- |
| `transformers` not found | Activate your venv. |
| Torch install fails | Use CPU-only wheel. |
| Models download slowly | Cached after first use. |
| Encoding issues | Ensure UTF-8 terminal. |
| Issue | Cause / Solution | ### API Issues
| ---------------------------- | -------------------------------------------- |
| **`transformers` not found** | Check virtual environment activation. |
| **Torch fails to install** | Install CPU-only version from PyTorch index. |
| **Models download slowly** | Hugging Face caches them after first run. |
| **Unicode / accents broken** | Ensure terminal encoding is UTF‑8. |
### API-Specific Issues | Issue | Solution |
| -------------------- | --------------------------------------- |
| Issue | Cause / Solution | | `FastAPI` missing | `pip install fastapi uvicorn[standard]` |
| ----------------------------- | ----------------------------------------------------- | | Port in use | Change with `--port 8001` |
| **`FastAPI` not found** | Install with `pip install fastapi uvicorn[standard]`. | | CORS error | Edit `allow_origins` in `api/config.py` |
| **Port already in use** | Use `--port 8001` or kill process on port 8000. | | Validation error 422 | Check request body |
| **CORS errors in browser** | Check `allow_origins` in `src/api/config.py`. | | 500 error | Verify model loading |
| **422 Validation Error** | Check request body matches Pydantic models. |
| **500 Internal Server Error** | Check model loading and pipeline initialization. |
### Quick API Health Check
```bash
# Test if API is running
curl http://localhost:8000/health
# Test basic endpoint
curl -X POST "http://localhost:8000/sentiment" \
-H "Content-Type: application/json" \
-d '{"text": "test"}'
```
--- ---
## 🧭 Development Guidelines ## 🧭 Development Guidelines
- Keep **Command** classes lightweight — no ML logic inside them. - Keep command classes lightweight (no ML inside)
- Reuse the **Pipeline Template** for new experiments. - Use the pipeline template for new tasks
- Format outputs consistently via the `DisplayFormatter`. - Format all outputs via `DisplayFormatter`
- Document all new models or commands in `README.md` and `settings.py`. - Document new commands and models
--- ---
## 🧱 Roadmap ## 🧱 Roadmap
- [ ] Add non-interactive CLI flags (`--text`, `--task`) - [ ] Non-interactive CLI flags (`--text`, `--task`)
- [ ] Add multilingual model options - [ ] Multilingual models
- [ ] Add automatic test coverage - [ ] Test coverage
- [ ] Add logging and profiling utilities - [ ] Logging & profiling
- [ ] Add export to JSON/CSV results - [ ] Export to JSON/CSV
--- ---
## 📜 License ## 📜 License
This project is licensed under the [MIT License](./LICENSE) — feel free to use it, modify it, and share it! Licensed under the [MIT License](./LICENSE).
You are free to use, modify, and distribute this project.
--- ---
✨ **End of Documentation**
_The AI Lab – Transformers CLI Playground: built for learning, experimenting, and sharing NLP excellence._

2
src/api/app.py
View File

@ -10,7 +10,7 @@ import logging
from .models import ( from .models import (
TextRequest, TextListRequest, QARequest, FillMaskRequest, TextRequest, TextListRequest, QARequest, FillMaskRequest,
SentimentResponse, NERResponse, QAResponse, FillMaskResponse, SentimentResponse, NERResponse, QAResponse, FillMaskResponse,
ModerationResponse, TextGenResponse, BatchResponse, ErrorResponse ModerationResponse, TextGenResponse, BatchResponse
) )
# Global pipeline instances # Global pipeline instances

17
src/api/config.py
View File

@ -2,6 +2,7 @@
API configuration settings API configuration settings
""" """
from typing import Dict, Any from typing import Dict, Any
from src.config.settings import Config
class APIConfig: class APIConfig:
@ -25,21 +26,6 @@ class APIConfig:
MAX_TEXT_LENGTH = 10000 MAX_TEXT_LENGTH = 10000
MAX_BATCH_SIZE = 100 MAX_BATCH_SIZE = 100
# Model defaults
DEFAULT_MODELS: Dict[str, str] = {
"sentiment": "cardiffnlp/twitter-roberta-base-sentiment-latest",
"ner": "dbmdz/bert-large-cased-finetuned-conll03-english",
"qa": "distilbert-base-cased-distilled-squad",
"fillmask": "bert-base-uncased",
"moderation": "martin-ha/toxic-comment-model",
"textgen": "gpt2"
}
@classmethod
def get_default_model(cls, task: str) -> str:
"""Get default model for a task"""
return cls.DEFAULT_MODELS.get(task, "")
@classmethod @classmethod
def get_all_settings(cls) -> Dict[str, Any]: def get_all_settings(cls) -> Dict[str, Any]:
"""Get all configuration settings""" """Get all configuration settings"""
@ -57,5 +43,4 @@ class APIConfig:
"max_text_length": cls.MAX_TEXT_LENGTH, "max_text_length": cls.MAX_TEXT_LENGTH,
"max_batch_size": cls.MAX_BATCH_SIZE "max_batch_size": cls.MAX_BATCH_SIZE
}, },
"default_models": cls.DEFAULT_MODELS
} }

10
src/config/settings.py
View File

@ -15,11 +15,11 @@ class Config:
# Default models # Default models
DEFAULT_MODELS = { DEFAULT_MODELS = {
"sentiment": "cardiffnlp/twitter-roberta-base-sentiment-latest", "sentiment": "distilbert-base-uncased-finetuned-sst-2-english",
"fillmask": "distilbert-base-uncased", "fillmask": "bert-base-uncased",
"textgen": "gpt2", "textgen": "gpt2",
"moderation": "unitary/toxic-bert", "ner": "dslim/bert-base-NER",
"ner": "dbmdz/bert-large-cased-finetuned-conll03-english", "moderation":"unitary/toxic-bert",
"qa": "distilbert-base-cased-distilled-squad", "qa": "distilbert-base-cased-distilled-squad",
} }