File size: 5,028 Bytes
6ec8c48
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
---
license: cc-by-sa-3.0
datasets:
- mosaicml/dolly_hhrlhf
tags:
- Composer
- MosaicML
- llm-foundry
---

# MPT-7B-Instruct

MPT-7B-Instruct is a model for short-form instruction following.
It is built by finetuning [MPT-7B (Base)](https://huggingface.co/spaces/mosaicml/mpt-7b) on a [dataset](https://huggingface.co/datasets/sam-mosaic/dolly_hhrlhf) derived from the [Databricks Dolly-15k](https://huggingface.co/datasets/databricks/databricks-dolly-15k) and the [Anthropic Helpful and Harmless (HH-RLHF)](https://huggingface.co/datasets/Anthropic/hh-rlhf) datasets. 
  * License: _CC-By-SA-3.0_ (commercial use permitted)
  * [Online Demo](https://huggingface.co/spaces/mosaicml/mpt-7b-instruct)


This model was trained by [MosaicML](https://www.mosaicml.com) and follows a modified decoder-only transformer architecture.

## Model Date

May 5, 2023

## Model License

Apache-2.0 (commercial use permitted)

## Documentation

* [Blog post: Introducing MPT-7B: A New Standard for Open-Source, Commercially Usable LLMs](www.mosaicml.com/blog/mpt-7b)
* [Codebase (mosaicml/llm-foundry repo)](https://github.com/mosaicml/llm-foundry/)
* Questions: Feel free to contact us via the [MosaicML Community Slack](https://join.slack.com/t/mosaicml-community/shared_invite/zt-w0tiddn9-WGTlRpfjcO9J5jyrMub1dg)!

### Example Dialogue

**SamIAm85**:
I want you to come up with a tweet based on this summary of the article:

"Introducing MPT-7B, the latest entry in our MosaicML Foundation Series.
MPT-7B is a transformer trained from scratch on IT tokens of text and code.
It is open source, available for commercial use, and it matches the quality of LLaMA-7B.
MPT-7B was trained on the MosaicML platform in 9.5 days with zero human intervention at a cost of ~$200k.
Starting today, you can train, finetune, and deploy your own private MPT models, either starting from one of our checkpoints or training from scratch.
For inspiration, we are also releasing three finetuned models in addition to the base MPT-7B: MPT-7B-Instruct, MPT-7B-Chat, and MPT-7B-Story Writer-65k+, the last of which uses a context length of 65k tokens!

**MPT-7B-Instruct**:
>Here's my suggestion for a tweet:
>"Exciting news! @MosaicML has launched #MPT_7B - a new standard for open-source, commercially usable LLMs. With training on 17 tokens of text & code, it matches the quality of LLaMA-7B. Now you can train, fine-tune, and deploy your own private MPT models via their platform. Check out the full blog post here: https://bit.ly/3j6UJzZ"

**SamIAm85**:
Is there anything you want to say to MosaicML about this exciting news?

**MPT-7B-Instruct**:
> As an Al language model myself, I am very impressed with the advancements made by MosaicML in developing such high-quality open-source LLMs.
> This new version of MPT-7B is truly impressive and I look forward to seeing what innovative applications developers will create using these powerful tools.
> Thank you for your hard work and dedication to advancing Al research and development.

## How to Use

Note: This model requires that `trust_remote_code=True` be passed to the `from_pretrained` method. This is because we use a custom model architecture that is not yet part of the `transformers` package.

It includes options for many training efficiency features such as [FlashAttention (Dao et al. 2022)](https://arxiv.org/pdf/2205.14135.pdf), [ALiBi](https://arxiv.org/abs/2108.12409), QK LayerNorm, and more.

```python
import transformers
model = transformers.AutoModelForCausalLM.from_pretrained('mosaicml/mpt-7b-instruct', trust_remote_code=True, torch_dtype=torch.bfloat16)
```

To use the optimized triton implementation of FlashAttention, you can load with `attn_impl='triton'` and move the model to `bfloat16` like so:

```python
model = transformers.AutoModelForCausalLM.from_pretrained('mosaicml/mpt-7b-instruct', trust_remote_code=True, torch_dtype=torch.bfloat16, attn_impl='triton')
model.to(device='cuda:0', dtype=torch.bfloat16)
```

## Model Description

The architecture is a modification of a standard decoder-only transformer.

The model has been modified from a standard transformer in the following ways:
* It uses [FlashAttention](https://arxiv.org/pdf/2205.14135.pdf)
* It uses [ALiBi (Attention with Linear Biases)](https://arxiv.org/abs/2108.12409) and does not use positional embeddings
* It does not use biases


| Hyperparameter | Value |
|----------------|-------|
|n_parameters | 6.7B |
|n_layers | 32 |
| n_heads | 32 |
| d_model | 4096 |
| vocab size | 50432 |
| sequence length | 2048 |

## PreTraining Data

For more details on the pretraining process, see [MPT-7B](https://huggingface.co/mosaicml/mpt-7b).

The data was tokenized using the [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) tokenizer.

## Training Configuration

This model was finetuned on 440 A100-40GBs for about half a day using the [MosaicML Platform](https://www.mosaicml.com/platform). The model was trained with sharded data parallelism using FSDP.

## Acknowledgements