ColossalAI/applications/Chat/examples/README.md

# Examples

## Table of Contents

- [Examples](#examples)
  - [Table of Contents](#table-of-contents)
  - [Install requirements](#install-requirements)
  - [Supervised datasets collection](#supervised-datasets-collection)
    - [Conversation dataset generation](#conversation-dataset-generation)
  - [Stage1 - Supervised instructs tuning](#stage1---supervised-instructs-tuning)
    - [Arg List](#arg-list)
  - [Stage2 - Training reward model](#stage2---training-reward-model)
    - [Features and tricks in RM training](#features-and-tricks-in-rm-training)
    - [Experiment result](#experiment-result)
    - [Arg List](#arg-list-1)
  - [Stage3 - Training model using prompts with RL](#stage3---training-model-using-prompts-with-rl)
    - [Arg List](#arg-list-2)
  - [Inference example - After Stage3](#inference-example---after-stage3)
  - [Attention](#attention)
    - [data](#data)
  - [Support Model](#support-model)
    - [GPT](#gpt)
    - [BLOOM](#bloom)
    - [OPT](#opt)
    - [LLaMA](#llama)
  - [Add your own models](#add-your-own-models)
    - [Actor model](#actor-model)
    - [Reward model](#reward-model)
    - [Critic model](#critic-model)

---

## Install requirements

```shell
pip install -r requirements.txt
```

## Supervised datasets collection

We collected 104K bilingual datasets of Chinese and English, and you can find the datasets in this repo
[InstructionWild](https://github.com/XueFuzhao/InstructionWild) and in this [file](https://github.com/XueFuzhao/InstructionWild/blob/main/data/README.md).

Here is how we collected the data

<p align="center">
<img src="https://raw.githubusercontent.com/hpcaitech/public_assets/main/applications/chat/data-collect.png" width=500/>
</p>

### Conversation dataset generation

In order to further improve the model's ability to handle multi-turn conversations, we need to include samples with multi-turn conversations in the dataset. However, the samples in InstructWild and Alpaca datasets currently consist of only single-turn conversations, and their dataset organization is not suitable for storing multi-turn conversations. Additionally, after converting the aforementioned datasets, we also need to include multi-turn conversation datasets like ShareGPT, and we should transform them into the training format supported by ColossalChat.

A sample of conversation dataset should have the following fields:

- `type` (str, optional): The type of the data sample.
- `language` (str, optional): The language of the data sample.
- `dataset` (str, optional): The dataset the data sample originates from.
- `conversations` (str, compulsory): Conversation content of the data sample.
- `id` (int, optional): The ID of the data sample.

A simple example:

```json
{
  "type": "instruction",
  "language": "English",
  "dataset": "Alpaca",
  "conversations": [
    {
      "from": "human",
      "value": "Give three tips for staying healthy."
    },
    {
      "from": "gpt",
      "value": "1.Eat a balanced diet and make sure to include plenty of fruits and vegetables. \n2. Exercise regularly to keep your body active and strong. \n3. Get enough sleep and maintain a consistent sleep schedule."
    }
  ],
  "id": 1
}
```

> **NOTE:** Only key `conversations` is compulsary for training and other keys serve as metadata. The length of `conversations` varies.

You can run the `examples/generate_conversation_dataset.py` to generate a conversation dataset supported by ColossalChat.

You can use the following cmd to generate conversation dataset.

```bash
python generate_conversation_dataset.py \
    --dataset "All"
    --save_path "/path/to/dataset"
```

## Stage1 - Supervised instructs tuning

Stage1 is supervised instructs fine-tuning, which uses the datasets mentioned earlier to fine-tune the model.
[[Stage1 tutorial video]](https://www.youtube.com/watch?v=-qFBZFmOJfg)

You can run the `examples/train_sft.sh` to start a supervised instructs fine-tuning.

You can also use the following cmd to start a supervised instructs fine-tuning with your own settings.

```bash
torchrun --standalone --nproc_per_node=4 train_sft.py \
    --pretrain "/path/to/LLaMa-7B/" \
    --model 'llama' \
    --strategy colossalai_zero2 \
    --save_path  /path/to/Coati-7B \
    --dataset /path/to/data.json \
    --batch_size 4 \
    --accumulation_steps 8 \
    --lr 2e-5 \
    --max_datasets_size 512 \
    --max_epochs 1 \
    --grad_checkpoint
```

**Note**: the supervised dataset follows the following format,

```json
[
    {
        "instruction": "Provide a list of the top 10 most popular mobile games in Asia",
        "input": "",
        "output": "The top 10 most popular mobile games in Asia are:\n1) PUBG Mobile\n2) Pokemon Go\n3) Candy Crush Saga\n4) Free Fire\n5) Clash of Clans\n6) Mario Kart Tour\n7) Arena of Valor\n8) Fantasy Westward Journey\n9) Subway Surfers\n10) ARK Survival Evolved",
        "id": 0
    },
    ...
]
```

### Arg List

- `--strategy`: the strategy using for training, choices=['ddp', 'colossalai_gemini', 'colossalai_zero2'], default='colossalai_zero2'
- `--model`: model type, choices=['gpt2', 'bloom', 'opt', 'llama'], default='bloom'
- `--pretrain`: pretrain model, type=str, default=None
- `--max_datasets_size`: the max size of dataset, type=int, default=None
- `--save_path`: path to save the model, type=str, default='output'
- `--need_optim_ckpt`: whether to save optim ckpt, type=bool, default=False
- `--max_epochs`: max epochs for training, type=int, default=3
- `--batch_size`: batch size while training, type=int, default=4
- `--lora_rank`: low-rank adaptation matrices rank, type=int, default=0
- `--grad_checkpoint`: enable gradient checkpointing, type=bool, default=False

## Stage2 - Training reward model

We train a reward model in stage 2, which obtains corresponding scores by manually ranking different outputs for the same prompt and supervises the training of the reward model.
[[Stage2 tutorial video]](https://www.youtube.com/watch?v=gMx2CApKhuo)

You can run the `examples/train_rm.sh` to start a reward model training.

You can also use the following cmd to start training a reward model.

```bash
torchrun --standalone --nproc_per_node=4 train_reward_model.py \
    --pretrain "/path/to/LLaMa-7B/" \
    --model 'llama' \
    --strategy colossalai_zero2 \
    --loss_fn 'log_exp'\
    --save_path 'rmstatic.pt' \
```

### Features and tricks in RM training

- We support [Anthropic/hh-rlhf](https://huggingface.co/datasets/Anthropic/hh-rlhf)and[rm-static](https://huggingface.co/datasets/Dahoas/rm-static) datasets.
- We support 2 kinds of loss function named `log_sig`(used by OpenAI) and `log_exp`(used by Anthropic).
- We change the loss to `valid_acc` and `pair_dist` to monitor progress during training.
- We add special token to the end of the sequence to get better result.
- We use cosine-reducing lr-scheduler for RM training.
- We set value_head as 1 liner layer and initialize the weight of value_head using N(0，1/(d_model + 1)) distribution.
- We train a Bloom-560m reward model for 1 epoch and find the test acc of the model achieve the performance mentions in [Anthropics paper](https://arxiv.org/abs/2204.05862).

### Experiment result

Model performance in [Anthropics paper](https://arxiv.org/abs/2204.05862):

<div align=middle> <img width="512" alt="image" src="https://user-images.githubusercontent.com/70618399/225263321-8d64c3a8-6877-4cc8-9b61-0e1c52d3d94f.png">

<div align=left>Our training & test result of bloom-560m for 1 epoch:

<div align=middle> <img width="512" alt="image" src="https://user-images.githubusercontent.com/70618399/225262950-a7f0a686-25de-44ec-98f2-11b83ea86674.png">

<div align=left>We also train the reward model based on LLaMA-7B, which reaches the ACC of 72.06% after 1 epoch, performing almost the same as Anthropic's best RM.

### Arg List

- `--strategy`: the strategy using for training, choices=['ddp', 'colossalai_gemini', 'colossalai_zero2'], default='colossalai_zero2'
- `--model`: model type, choices=['gpt2', 'bloom', 'opt', 'llama'], default='bloom'
- `--pretrain`: pretrain model, type=str, default=None
- `--model_path`: the path of rm model(if continue to train), type=str, default=None
- `--save_path`: path to save the model, type=str, default='output'
- `--need_optim_ckpt`: whether to save optim ckpt, type=bool, default=False
- `--max_epochs`: max epochs for training, type=int, default=3
- `--dataset`: dataset name, type=str, choices=['Anthropic/hh-rlhf', 'Dahoas/rm-static']
- `--subset`: subset of the dataset, type=str, default=None
- `--batch_size`: batch size while training, type=int, default=4
- `--lora_rank`: low-rank adaptation matrices rank, type=int, default=0
- `--loss_func`: which kind of loss function, choices=['log_sig', 'log_exp']
- `--max_len`: max sentence length for generation, type=int, default=512

## Stage3 - Training model using prompts with RL

Stage3 uses reinforcement learning algorithm, which is the most complex part of the training process, as shown below:

<p align="center">
<img src="https://raw.githubusercontent.com/hpcaitech/public_assets/main/applications/chat/stage-3.jpeg" width=800/>
</p>

You can run the `examples/train_prompts.sh` to start PPO training.

You can also use the cmd following to start PPO training.
[[Stage3 tutorial video]](https://www.youtube.com/watch?v=Z8wwSHxPL9g)

```bash
torchrun --standalone --nproc_per_node=4 train_prompts.py \
    --pretrain "/path/to/LLaMa-7B/" \
    --model 'llama' \
    --strategy colossalai_zero2 \
    --prompt_dataset /path/to/your/prompt_dataset \
    --pretrain_dataset /path/to/your/pretrain_dataset \
    --rm_pretrain /your/pretrain/rm/definition \
    --rm_path /your/rm/model/path
```

Prompt dataset: the instruction dataset mentioned in the above figure which includes the instructions, e.g. you can use the [script](https://github.com/hpcaitech/ColossalAI/tree/main/applications/Chat/examples/generate_prompt_dataset.py) which samples `instinwild_en.json` or `instinwild_ch.json` in [InstructionWild](https://github.com/XueFuzhao/InstructionWild/tree/main/data#instructwild-data) to generate the prompt dataset.
Pretrain dataset: the pretrain dataset including the instruction and corresponding response, e.g. you can use the [InstructWild Data](https://github.com/XueFuzhao/InstructionWild/tree/main/data) in stage 1 supervised instructs tuning.

**Note**: the required datasets follow the following format,

- `pretrain dataset`

  ```json
  [
      {
          "instruction": "Provide a list of the top 10 most popular mobile games in Asia",
          "input": "",
          "output": "The top 10 most popular mobile games in Asia are:\n1) PUBG Mobile\n2) Pokemon Go\n3) Candy Crush Saga\n4) Free Fire\n5) Clash of Clans\n6) Mario Kart Tour\n7) Arena of Valor\n8) Fantasy Westward Journey\n9) Subway Surfers\n10) ARK Survival Evolved",
          "id": 0
      },
      ...
  ]
  ```

- `prompt dataset`

  ```json
  [
      {
          "instruction": "Edit this paragraph to make it more concise: \"Yesterday, I went to the store and bought some things. Then, I came home and put them away. After that, I went for a walk and met some friends.\"",
          "id": 0
      },
      {
          "instruction": "Write a descriptive paragraph about a memorable vacation you went on",
          "id": 1
      },
      ...
  ]
  ```

### Arg List

- `--strategy`: the strategy using for training, choices=['ddp', 'colossalai_gemini', 'colossalai_zero2'], default='colossalai_zero2'
- `--model`: model type of actor, choices=['gpt2', 'bloom', 'opt', 'llama'], default='bloom'
- `--pretrain`: pretrain model, type=str, default=None
- `--rm_model`: reward model type, type=str, choices=['gpt2', 'bloom', 'opt', 'llama'], default=None
- `--rm_pretrain`: pretrain model for reward model, type=str, default=None
- `--rm_path`: the path of rm model, type=str, default=None
- `--save_path`: path to save the model, type=str, default='output'
- `--prompt_dataset`: path of the prompt dataset, type=str, default=None
- `--pretrain_dataset`: path of the ptx dataset, type=str, default=None
- `--need_optim_ckpt`: whether to save optim ckpt, type=bool, default=False
- `--num_episodes`: num of episodes for training, type=int, default=10
- `--num_update_steps`: number of steps to update policy per episode, type=int
- `--num_collect_steps`: number of steps to collect experience per episode, type=int
- `--train_batch_size`: batch size while training, type=int, default=8
- `--ptx_batch_size`: batch size to compute ptx loss, type=int, default=1
- `--experience_batch_size`: batch size to make experience, type=int, default=8
- `--lora_rank`: low-rank adaptation matrices rank, type=int, default=0
- `--kl_coef`: kl_coef using for computing reward, type=float, default=0.1
- `--ptx_coef`: ptx_coef using for computing policy loss, type=float, default=0.9

## Inference example - After Stage3

We support different inference options, including int8 and int4 quantization.
For details, see [`inference/`](https://github.com/hpcaitech/ColossalAI/tree/main/applications/Chat/inference).

## Attention

The examples are demos for the whole training process.You need to change the hyper-parameters to reach great performance.

#### data

- [x] [rm-static](https://huggingface.co/datasets/Dahoas/rm-static)
- [x] [hh-rlhf](https://huggingface.co/datasets/Anthropic/hh-rlhf)
- [ ] [openai/summarize_from_feedback](https://huggingface.co/datasets/openai/summarize_from_feedback)
- [ ] [openai/webgpt_comparisons](https://huggingface.co/datasets/openai/webgpt_comparisons)
- [ ] [Dahoas/instruct-synthetic-prompt-responses](https://huggingface.co/datasets/Dahoas/instruct-synthetic-prompt-responses)

## Support Model

### GPT

- [x] GPT2-S (s)
- [x] GPT2-M (m)
- [x] GPT2-L (l)
- [x] GPT2-XL (xl)
- [x] GPT2-4B (4b)
- [ ] GPT2-6B (6b)

### BLOOM

- [x] [BLOOM-560m](https://huggingface.co/bigscience/bloom-560m)
- [x] [BLOOM-1b1](https://huggingface.co/bigscience/bloom-1b1)
- [x] [BLOOM-3b](https://huggingface.co/bigscience/bloom-3b)
- [x] [BLOOM-7b](https://huggingface.co/bigscience/bloom-7b1)
- [ ] [BLOOM-175b](https://huggingface.co/bigscience/bloom)

### OPT

- [x] [OPT-125M](https://huggingface.co/facebook/opt-125m)
- [x] [OPT-350M](https://huggingface.co/facebook/opt-350m)
- [x] [OPT-1.3B](https://huggingface.co/facebook/opt-1.3b)
- [x] [OPT-2.7B](https://huggingface.co/facebook/opt-2.7b)
- [x] [OPT-6.7B](https://huggingface.co/facebook/opt-6.7b)
- [ ] [OPT-13B](https://huggingface.co/facebook/opt-13b)
- [ ] [OPT-30B](https://huggingface.co/facebook/opt-30b)

### [LLaMA](https://github.com/facebookresearch/llama/blob/main/MODEL_CARD.md)

- [x] LLaMA-7B
- [x] LLaMA-13B
- [ ] LLaMA-33B
- [ ] LLaMA-65B

## Add your own models

If you want to support your own model in Coati, please refer the pull request for RoBERTa support as an example --[[chatgpt] add pre-trained model RoBERTa for RLHF stage 2 & 3](https://github.com/hpcaitech/ColossalAI/pull/3223), and submit a PR to us.

You should complete the implementation of four model classes, including Reward model, Critic model, LM model, Actor model

here are some example code for a NewModel named `Coati`.
if it is supported in huggingface [transformers](https://github.com/huggingface/transformers), you can load it by `from_pretrained`, o
r you can build your own model by yourself.

### Actor model

```python
from ..base import Actor
from transformers.models.coati import CoatiModel

class CoatiActor(Actor):
    def __init__(self,
                 pretrained: Optional[str] = None,
                 checkpoint: bool = False,
                 lora_rank: int = 0,
                 lora_train_bias: str = 'none') -> None:
        if pretrained is not None:
            model = CoatiModel.from_pretrained(pretrained)
        else:
            model = build_model() # load your own model if it is not support in transformers

        super().__init__(model, lora_rank, lora_train_bias)
```

### Reward model

```python
from ..base import RewardModel
from transformers.models.coati import CoatiModel

class CoatiRM(RewardModel):

    def __init__(self,
                 pretrained: Optional[str] = None,
                 checkpoint: bool = False,
                 lora_rank: int = 0,
                 lora_train_bias: str = 'none') -> None:
        if pretrained is not None:
            model = CoatiModel.from_pretrained(pretrained)
        else:
            model = build_model() # load your own model if it is not support in transformers

        value_head = nn.Linear(model.config.n_embd, 1)
        value_head.weight.data.normal_(mean=0.0, std=1 / (model.config.n_embd + 1))
        super().__init__(model, value_head, lora_rank, lora_train_bias)
```

### Critic model

```python
from ..base import Critic
from transformers.models.coati import CoatiModel

class CoatiCritic(Critic):
    def __init__(self,
                 pretrained: Optional[str] = None,
                 checkpoint: bool = False,
                 lora_rank: int = 0,
                 lora_train_bias: str = 'none') -> None:
        if pretrained is not None:
            model = CoatiModel.from_pretrained(pretrained)
        else:
            model = build_model() # load your own model if it is not support in transformers

        value_head = nn.Linear(model.config.n_embd, 1)
        value_head.weight.data.normal_(mean=0.0, std=1 / (model.config.n_embd + 1))
        super().__init__(model, value_head, lora_rank, lora_train_bias)
```