File size: 18,213 Bytes
4f319ce
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d543ee8
4f319ce
 
 
 
 
 
 
 
1402e8d
4f319ce
 
 
 
cfa9881
4f319ce
 
 
 
 
 
 
 
 
 
 
 
 
 
d543ee8
 
 
 
 
 
4f319ce
 
 
d543ee8
4f319ce
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
---
license: apache-2.0
tags:
- OpenAccess AI Collective
- MPT
- axolotl
datasets:
- ehartford/WizardLM_alpaca_evol_instruct_70k_unfiltered
- QingyiSi/Alpaca-CoT
- teknium/GPTeacher-General-Instruct
- metaeval/ScienceQA_text_only
- hellaswag
- openai/summarize_from_feedback
- riddle_sense
- gsm8k
- camel-ai/math
- camel-ai/biology
- camel-ai/physics
- camel-ai/chemistry
- winglian/evals

inference: false
---

[<img src="https://raw.githubusercontent.com/OpenAccess-AI-Collective/axolotl/main/image/axolotl-badge-web.png" alt="Built with Axolotl" width="200" height="32"/>](https://github.com/OpenAccess-AI-Collective/axolotl)

# Minotaur MPT 7B

Minotaur 7B is an instruct fine-tuned model on top of MPT-7B. Minotaur 7B is fine-tuned **on only completely open datasets** making this model reproducible by anyone.

Questions, comments, feedback, looking to donate, or want to help? Reach out on our [Discord](https://discord.gg/PugNNHAF5r) or email [wing@openaccessaicollective.org](mailto:wing@openaccessaicollective.org)

# Prompts
Chat only style prompts using `USER:`,`ASSISTANT:`.

# Training Datasets

Minotaur 7B model is fine-tuned on the following datasets:

- [WizardLM](https://huggingface.co/datasets/ehartford/WizardLM_alpaca_evol_instruct_70k_unfiltered)
- [subset of QingyiSi/Alpaca-CoT for roleplay and CoT](https://huggingface.co/QingyiSi/Alpaca-CoT)
- [GPTeacher-General-Instruct](https://huggingface.co/datasets/teknium/GPTeacher-General-Instruct)
- [metaeval/ScienceQA_text_only](https://huggingface.co/datasets/metaeval/ScienceQA_text_only) - instruct for concise responses
- [openai/summarize_from_feedback](https://huggingface.co/datasets/openai/summarize_from_feedback) - instruct augmented tl;dr summarization
- [camel-ai/math](https://huggingface.co/datasets/camel-ai/math)
- [camel-ai/physics](https://huggingface.co/datasets/camel-ai/physics)
- [camel-ai/chemistry](https://huggingface.co/datasets/camel-ai/chemistry)
- [camel-ai/biology](https://huggingface.co/datasets/camel-ai/biology)
- [winglian/evals](https://huggingface.co/datasets/winglian/evals)
  - custom sysnthetic datasets around misconceptions, in-context qa, jokes, N-tasks problems, and context-insensitivity
  - ARC-Easy & ARC-Challenge - instruct augmented for detailed responses, derived from the `train` split
  - [hellaswag](https://huggingface.co/datasets/hellaswag) - 30K+ rows of instruct augmented for detailed explanations w 30K+ rows, derived from the `train` split
  - [riddle_sense](https://huggingface.co/datasets/riddle_sense) - instruct augmented 
  - [gsm8k](https://huggingface.co/datasets/gsm8k) - instruct augmented 

# Shoutouts

Special thanks to Nanobit for helping with Axolotl and TheBloke for quantizing these models are more accessible to all.

# Demo

HF Demo in Spaces coming soon.

## Release Notes

- https://wandb.ai/wing-lian/mpt-7b-4k-minotaur/runs/i4zib0j4

## Build

Minotaur was built with [Axolotl](https://github.com/OpenAccess-AI-Collective/axolotl) on 8xA100 80GB 
 - 3 epochs taking approximately 6 hours

## Bias, Risks, and Limitations
Minotaur has not been aligned to human preferences with techniques like RLHF or deployed with in-the-loop filtering of responses like ChatGPT, so the model can produce problematic outputs (especially when prompted to do so).
Minotaur was fine-tuned from the base model MPT-7B, please refer to its model card's Limitations Section for relevant information. (included below)

## Examples - results may vary based on temperature and other settings

```
USER: your have 3 apples. you eat 2 pears. how many apples do you have left?

ASSISTANT:To find the quantity of apples, we can subtract the number of pears from the number of apples.

In this case, there are 3 apples and 2 pears. This means there are 3 apples left.
```
wut?


```
USER: your have 3 apples. you eat 2 pears. how many apples do you have left?

ASSISTANT:You don't have any apples left. You had 3 and after eating 2 pears, it is 2. You ate the other apple.
```
🤣


```
USER: Solve for x: 5x = 3x + 10

ASSISTANT:Solve for x: 5x = 3x + 10

We can subtract 3x from each side of the equation:
5(x - 3) = 10
x = 5 + 3
x = 8
```
yup, it still sucks at math

```
USER: tell me a joke. then explain why it's funny.

ASSISTANT: What did the blind man say to the man eating steak?
He said, "Is that a cow or a sandwich?"
```

```
USER: write a sonnet about doves

ASSISTANT:Doves are a beautiful bird, and you can write a sonnet about them to express your love for them. This sonnet can include imagery of doves, including the sound of their wings as they fly, the softness of their feathers, and the peacefulness they bring to their surroundings. The structure of this sonnet can be similar to the Shakespearean sonnets, with 14 lines divided into three groups of rhyming couplets. In the first stanza, you can express your love for doves as they are a symbol of peace and tranquility. The second stanza can describe the beauty of the dove's plumage and the softness of their feathers. The third stanza can be about the peacefulness that they bring to their surroundings, such as in a garden or park. You can end the sonnet with a final line that encapsulates your feelings about doves.

Here is the sonnet for your reference:

> Doves are a symbol of peace and tranquility,
> A gentle song that soothes the soul.
> Their soft feathers and sweet voices bring a sense of warmth and calm to their surroundings,
> And they are a welcome sight in any landscape.
> 
> Their grace and beauty are a reminder of the simple pleasures in life,
> And their peaceful presence is a welcome respite from the stresses of everyday life.
> 
> Doves are a symbol of love and friendship,
> And their companionship is a gift we can all cherish.
> 
> Let us embrace the beauty and serenity of doves,
> And let us cherish the gift of their presence in our lives.
> 
> May we always find beauty and serenity in the simple pleasures of life,
> And may we cherish the gift of friendship and love.
> 
> Doves are a symbol of peace and tranquility,
> A gentle song that soothes the soul.
> Their soft feathers and sweet voices bring a sense of warmth and calm to their surroundings,
> And they are a welcome sight in any landscape.
> 
> Their grace and beauty are a reminder of the simple pleasures in life,
> And their peaceful presence is a welcome respite from the stresses of everyday life.
> 
> Doves are a symbol of love and friendship,
> And their companionship is a gift we can all cherish.
> 
> Let us embrace the beauty and serenity of doves,
> And let us cherish the gift of their presence in our lives.
> 
> May we always find beauty and serenity in the simple pleasures of life,
> And may we cherish the gift of friendship and love.
```

# MPT-7B

MPT-7B is a decoder-style transformer pretrained from scratch on 1T tokens of English text and code.
This model was trained by [MosaicML](https://www.mosaicml.com).

MPT-7B is part of the family of MosaicPretrainedTransformer (MPT) models, which use a modified transformer architecture optimized for efficient training and inference.

These architectural changes include performance-optimized layer implementations and the elimination of context length limits by replacing
positional embeddings with Attention with Linear Biases ([ALiBi](https://arxiv.org/abs/2108.12409)).
Thanks to these modifications, MPT models can be trained with high throughput efficiency and stable convergence.
MPT models can also be served efficiently with both standard HuggingFace pipelines and NVIDIA's [FasterTransformer](https://github.com/NVIDIA/FasterTransformer).

This model uses the MosaicML LLM codebase, which can be found in the [llm-foundry repository](https://github.com/mosaicml/llm-foundry). It was trained by MosaicML’s NLP team on the [MosaicML platform](https://www.mosaicml.com/training) for LLM pretraining, finetuning, and inference.

### How is this model different?

MPT-7B is

* **Licensed for the possibility of commercial use** (unlike [LLaMA](https://arxiv.org/abs/2302.13971)).
* **Trained on a large amount of data** (1T tokens like [LLaMA](https://arxiv.org/abs/2302.13971) vs. 300B for [Pythia](https://github.com/EleutherAI/pythia), 300B for [OpenLLaMA](https://github.com/openlm-research/open_llama), and 800B for [StableLM](https://github.com/Stability-AI/StableLM)).
* **Prepared to handle extremely long inputs** thanks to [ALiBi](https://arxiv.org/abs/2108.12409) (we finetuned [MPT-7B-StoryWriter-65k+](https://huggingface.co/mosaicml/mpt-7b-storywriter) on up to 65k inputs and can handle up to 84k vs. 2k-4k for other open source models).
* **Capable of fast training and inference** (via [FlashAttention](https://arxiv.org/pdf/2205.14135.pdf) and [FasterTransformer](https://github.com/NVIDIA/FasterTransformer))
* **Equipped with highly efficient open-source training code** via the [llm-foundry repository](https://github.com/mosaicml/llm-foundry)

### Models finetuned off MPT-7B:

The following models are finetuned on MPT-7B:

* [MPT-7B-StoryWriter-65k+](https://huggingface.co/mosaicml/mpt-7b-storywriter): a model designed to read and write fictional stories with super long context lengths.
Built by finetuning MPT-7B with a context length of 65k tokens on a filtered fiction subset of the [books3 dataset](https://huggingface.co/datasets/the_pile_books3).
At inference time, thanks to [ALiBi](https://arxiv.org/abs/2108.12409), MPT-7B-StoryWriter-65k+ can extrapolate even beyond 65k tokens.
We demonstrate generations as long as 80k tokens on a single A100-80GB GPU in our [blogpost](www.mosaicml.com/blog/mpt-7b).
  * License: Apache 2.0

* [MPT-7B-Instruct](https://huggingface.co/mosaicml/mpt-7b-instruct): a model for short-form instruction following.
Built by finetuning MPT-7B on a [dataset](https://huggingface.co/datasets/mosaicml/dolly_hhrlhf) we also release, 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_
  * [Demo on Hugging Face Spaces](https://huggingface.co/spaces/mosaicml/mpt-7b-instruct)

* [MPT-7B-Chat](https://huggingface.co/mosaicml/mpt-7b-chat): a chatbot-like model for dialogue generation.
Built by finetuning MPT-7B on the [ShareGPT-Vicuna](https://huggingface.co/datasets/jeffwan/sharegpt_vicuna), [HC3](https://huggingface.co/datasets/Hello-SimpleAI/HC3),
 [Alpaca](https://huggingface.co/datasets/tatsu-lab/alpaca), [HH-RLHF](https://huggingface.co/datasets/Anthropic/hh-rlhf), and [Evol-Instruct](https://huggingface.co/datasets/victor123/evol_instruct_70k) datasets.
  * License: _CC-By-NC-SA-4.0_
  * [Demo on Hugging Face Spaces](https://huggingface.co/spaces/mosaicml/mpt-7b-chat)

## Model Date

May 5, 2023

## Model License

Apache-2.0

## Documentation

* [Blog post: Introducing MPT-7B: A New Standard for Open-Source, Commercially Usable LLMs](https://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://mosaicml.me/slack)!


## How to Use

This model is best used with the MosaicML [llm-foundry repository](https://github.com/mosaicml/llm-foundry) for training and finetuning.

```python
import transformers
model = transformers.AutoModelForCausalLM.from_pretrained(
  'mosaicml/mpt-7b',
  trust_remote_code=True
)
```
Note: This model requires that `trust_remote_code=True` be passed to the `from_pretrained` method.
This is because we use a custom `MPT` model architecture that is not yet part of the Hugging Face `transformers` package.
`MPT` includes options for many training efficiency features such as [FlashAttention](https://arxiv.org/pdf/2205.14135.pdf), [ALiBi](https://arxiv.org/abs/2108.12409), [QK LayerNorm](https://arxiv.org/abs/2010.04245), and more.

To use the optimized [triton implementation](https://github.com/openai/triton) of FlashAttention, you can load the model on GPU (`cuda:0`) with `attn_impl='triton'` and with `bfloat16` precision:
```python
import torch
import transformers

name = 'mosaicml/mpt-7b'

config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True)
config.attn_config['attn_impl'] = 'triton'
config.init_device = 'cuda:0' # For fast initialization directly on GPU!

model = transformers.AutoModelForCausalLM.from_pretrained(
  name,
  config=config,
  torch_dtype=torch.bfloat16, # Load model weights in bfloat16
  trust_remote_code=True
)
```

Although the model was trained with a sequence length of 2048, ALiBi enables users to increase the maximum sequence length during finetuning and/or inference. For example:

```python
import transformers

name = 'mosaicml/mpt-7b'

config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True)
config.max_seq_len = 4096 # (input + output) tokens can now be up to 4096

model = transformers.AutoModelForCausalLM.from_pretrained(
  name,
  config=config,
  trust_remote_code=True
)
```

This model was trained with the [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) tokenizer.

```python
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained('EleutherAI/gpt-neox-20b')
```

## 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 |



## Training Data

### Streaming Datasets

Data was formatted using the MosaicML [StreamingDataset](https://github.com/mosaicml/streaming) library to host our data in object storage and efficiently stream it to our compute cluster during training.
StreamingDataset obviates the need to download the whole dataset before starting training, and allows instant resumption of training from any point in the dataset.


### Data Mix

The model was trained for 1T tokens (with batch size 1760 and sequence length 2048). It was trained on the following data mix:


| Data Source | Number of Tokens in Source | Proportion | Effective Number of Tokens | Epochs |
|-------------|----------------------------|------------|----------------------------|--------|
| mC4 3.1.0 - English | 417.99 B | 0.33 | 330 B | 0.14 |
| C4 - English - SemDedup 80% | 100.42 B | 0.299 | 299 B | 2.98 |
| RedPajama - CommonCrawl | 878.45 B | 0.1 | 100 B | 0.11 |
| The Stack - Selected Languages | 463.78 B | 0.1 | 100 B | 0.22 |
| RedPajama - Wikipedia - En | 4.87 B | 0.04 | 40 B | 8.21 |
| The Stack - Markdown | 107.07 B | 0.035 | 35 B | 0.33 |
| S2ORC | 48.85 B | 0.033 | 33 B | 0.68 |
| RedPajama - Books | 26.02 B | 0.03 | 30B | 1.15 |
| RedPajama - arXiv | 28.10 B | 0.019 | 19 B | 0.68 |
| RedPajama - StackExchange | 20.54 B | 0.014 | 14 B |0.68 |

Samples for each batch were selected from one of the datasets with the probability specified above.
The examples were shuffled within each dataset, and each example was constructed from as many sequences from that dataset as were necessary to fill the 2048 sequence length.

The data was tokenized using the [EleutherAI/gpt-neox-20b](https://huggingface.co/EleutherAI/gpt-neox-20b) tokenizer. This BPE tokenizer has a number of desirable characteristics,
most of which are relevant for tokenizing code:
(1) It was trained on a diverse mix of data that includes code (The Pile)
(2) It applies consistent space delimitation, unlike the GPT2 tokenizer which tokenizes inconsistently depending on the presence of prefix spaces
(3) It contains tokens for repeated space characters, which allows superior compression of text with large amounts of repeated space characters.

The model vocabulary size of 50432 was set to be a multiple of 128 (as in [MEGATRON-LM](https://arxiv.org/abs/1909.08053)), model flop utilization (MFU) increased by up to four percentage points.

### Training Configuration

This model was trained on 440 A100-40GBs for about 9.5 days using the [MosaicML Platform](https://www.mosaicml.com/platform).
The model was trained with sharded data parallelism using [FSDP](https://pytorch.org/docs/stable/fsdp.html) and used the [LION](https://arxiv.org/abs/2302.06675) optimizer.

## Limitations and Biases

_The following language is modified from [EleutherAI's GPT-NeoX-20B](https://huggingface.co/EleutherAI/gpt-neox-20b)_

MPT-7B (Base) is **not** intended for deployment without finetuning.
It should not be used for human-facing interactions without further guardrails and user consent.

MPT-7B can produce factually incorrect output, and should not be relied on to produce factually accurate information.
MPT-7B was trained on various public datasets.
While great efforts have been taken to clean the pretraining data, it is possible that this model could generate lewd, biased or otherwise offensive outputs.


## MosaicML Platform

If you're interested in [training](https://www.mosaicml.com/training) and [deploying](https://www.mosaicml.com/inference) your own MPT or LLMs on the MosaicML Platform, [sign up here](https://forms.mosaicml.com/demo?utm_source=huggingface&utm_medium=referral&utm_campaign=mpt-7b).

## Disclaimer

The license on this model does not constitute legal advice. We are not responsible for the actions of third parties who use this model. Please cosult an attorney before using this model for commercial purposes.

## Citation

Please cite this model using the following format:

```
@online{MosaicML2023Introducing,
    author    = {MosaicML NLP Team},
    title     = {Introducing MPT-7B: A New Standard for Open-Source,
    ly Usable LLMs},
    year      = {2023},
    url       = {www.mosaicml.com/blog/mpt-7b},
    note      = {Accessed: 2023-03-28}, % change this date
    urldate   = {2023-03-28} % change this date
}
```