# Dolly Generation

*author: Jael*

<br />

## Description

A LLM operator generates answer given prompt in messages using a large language model or service.
This operator uses a pretrained [Dolly](https://github.com/databrickslabs/dolly) to generate response.
It will download model from [HuggingFace Models](https://huggingface.co/models).

<br />

## Code Example

Use the default model to continue the conversation from given messages.

*Write a pipeline with explicit inputs/outputs name specifications:*

```python
from towhee import pipe, ops

p = (
    pipe.input('question', 'docs', 'history')
        .map(('question', 'docs', 'history'), 'prompt', ops.prompt.question_answer(llm_name='dolly'))
        .map('prompt', 'answer', ops.LLM.Dolly())
        .output('answer')
)

history=[('Who won the world series in 2020?', 'The Los Angeles Dodgers won the World Series in 2020.')]
question = 'Where was it played?'
answer = p(question, [], history).get()[0]
```

*Write a [retrieval-augmented generation pipeline](https://towhee.io/tasks/detail/pipeline/retrieval-augmented-generation) with explicit inputs/outputs name specifications:*

```python
from towhee import pipe, ops


temp = '''{question}

Input:
{context}
'''


docs = ['You can install Towhee via the command `pip install towhee`.']
history = [
    ('What is Towhee?', 'Towhee is an open-source machine learning project that helps you encode your unstructured data into embeddings.')
]
question = 'How to install it?'

p = (
    pipe.input('question', 'docs', 'history')
        .map(('question', 'docs', 'history'),
             'prompt',
             ops.prompt.template(temp, ['question', 'context']))
        .map('prompt', 'answer',
             ops.LLM.Dolly())
        .output('answer')
)

answer = p(question, docs, history).get()[0]
```

<br />

## Factory Constructor

Create the operator via the following factory method:

***LLM.Dolly(model_name: str)***

**Parameters:**

***model_name***: *str*

The model name in string, defaults to 'databricks/dolly-v2-12b'. Supported model names:
- databricks/dolly-v2-12b
- databricks/dolly-v2-7b
- databricks/dolly-v2-3b
- databricks/dolly-v1-6b

***\*\*kwargs***

Other Dolly model parameters such as device_map.

<br />

## Interface

The operator takes a piece of text in string as input.
It returns answer in json.

***\_\_call\_\_(txt)***

**Parameters:**

***messages***: *list*

​	A list of messages to set up chat.
Must be a list of dictionaries with key value from "system", "question", "answer". For example, [{"question": "a past question?", "answer": "a past answer."}, {"question": "current question?"}]

**Returns**:

*answer: str*

​	The answer generated.

<br />




# More Resources

- [Training Your Own Text Embedding Model - Zilliz blog](https://zilliz.com/learn/training-your-own-text-embedding-model): Explore how to train your text embedding model using the `sentence-transformers` library and generate our training data by leveraging a pre-trained LLM.
- [Boost your LLM with Private Data Using LlamaIndex | Zilliz Webinar](https://zilliz.com/event/boost-your-llm-with-private-data-using-llamaindex/success): Zilliz webinar covering how to boost your LLM with private data with LlamaIndex to generate accurate and meaningful responses that reflect unique data inputs.
- [LLama2 vs ChatGPT: How They Perform in Question Answering - Zilliz blog](https://zilliz.com/blog/comparing-meta-ai-Llama2-openai-chatgpt): What is Llama 2, and how does it perform in question answering compared to ChatGPT?