https://github.com/minggnim/nlp-models

A repository for training transformer based models
https://github.com/minggnim/nlp-models

chatbot chatbots ctransformers deeplearning falcon fine-tuning gpt-2 langchain llama2 llms multi-label-classification multi-task-learning nlp pytorch qdrant-vector-database transformers

Last synced: 25 days ago
JSON representation

A repository for training transformer based models

• Host: GitHub
• URL: https://github.com/minggnim/nlp-models
• Owner: minggnim
• License: mit
• Created: 2022-08-08T23:15:11.000Z (over 2 years ago)
• Default Branch: master
• Last Pushed: 2024-07-30T03:42:36.000Z (6 months ago)
• Last Synced: 2024-10-29T08:03:50.487Z (3 months ago)
• Topics: chatbot, chatbots, ctransformers, deeplearning, falcon, fine-tuning, gpt-2, langchain, llama2, llms, multi-label-classification, multi-task-learning, nlp, pytorch, qdrant-vector-database, transformers
• Language: Jupyter Notebook
• Homepage:
• Size: 15.9 MB
• Stars: 2
• Watchers: 2
• Forks: 1
• Open Issues: 1
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
![PyPI - Package Version](https://img.shields.io/pypi/v/nlp-models?logo=pypi&style=flat&color=blue)



[![PyPI pyversions](https://img.shields.io/pypi/pyversions/nlp-models.svg)](https://pypi.python.org/pypi/nlp-models/)

[![Python package](https://github.com/minggnim/nlp-classification-model/actions/workflows/python-package.yml/badge.svg)](https://github.com/minggnim/nlp-classification-model/actions/workflows/python-package.yml)

[![Dependency Review](https://github.com/minggnim/nlp-classification-model/actions/workflows/dependency-review.yml/badge.svg)](https://github.com/minggnim/nlp-classification-model/actions/workflows/dependency-review.yml)

# NLP Models

A repository for building transformer based nlp models

## Installation

### Install from PyPi

```

pip install nlp-models

```

### Install from source

```

git clone [email protected]:minggnim/nlp-models.git

pip install -r requirements

```

## Llama2 Quantization model on consumer CPU

### Run Chat applications on CPU

1. Streamlit UI

    ```

    cd apps

    streamlit run chat.py

    ```

2. Command line

    ```

    llm_app chat -s 'hi there'

    ```

### Run Q&A application on CPU

1. Steamlit UI

    ```

    cd apps

    streamlit run qa.py

    ```

## Models

1. `bert_classifier`

   A wrapper package around BERT-based classification models

   - [Training example](https://github.com/minggnim/nlp-models/blob/master/notebooks/01_bert-classification-finetuning/01_a_classification_model_training_example.ipynb)

   - [Inference example](https://github.com/minggnim/nlp-models/blob/master/notebooks/01_bert-classification-finetuning/01_b_classification_inference_example.ipynb)

2. `multi_task_model`

   An implementation of multi-tasking model built on encoder models

   - [Zero-shot multi-task model](https://github.com/minggnim/nlp-models/blob/master/notebooks/02_multi-task-model/02_a_multitask_model_zeroshot_learning.ipynb)

   - [Training example](https://github.com/minggnim/nlp-models/blob/master/notebooks/02_multi-task-model/02_b_multitask_model_training_example.ipynb)

   - [Inference example](https://github.com/minggnim/nlp-models/blob/master/notebooks/02_multi-task-model/02_c_multitask_model_inference_example.ipynb)

   - [Qqrant Vector DB](https://github.com/minggnim/nlp-models/blob/master/notebooks/02_multi-task-model/02_d_qdrant_vector_db.ipynb)

3. `GPT-2`

    - [Training GPT-2 model](https://github.com/minggnim/nlp-models/blob/master/notebooks/03_gpt-2-training/gpt-2-training/03_gpt2_training.ipynb)

4. `Falcon 7B`

    - [Running Falcon 7b model](https://github.com/minggnim/nlp-models/blob/master/notebooks/04_llms/05_falcon_7b.ipynb)

5. Quantized Llama2 models

    - [Run Llama2 chat on CPU](https://github.com/minggnim/nlp-models/blob/master/notebooks/04_llms/06_llama2_langchain_gglm_inference.ipynb)

    - [Run Llama2 QA on a custom pdf document on CPU](https://github.com/minggnim/nlp-models/blob/master/notebooks/04_llms/07_llama2_doc_qa.ipynb)