MosaicML's MPT-30B GGML
These files are GGML format model files for MosaicML's MPT-30B.
Please note that these GGMLs are not compatible with llama.cpp, or currently with text-generation-webui. Please see below for a list of tools known to work with these model files.
KoboldCpp just added GPU accelerated (OpenCL) support for MPT models, so that is the client I recommend using for these models.
Note: Please use version 1.32.1 or later of KoboldCpp, as there was a bug in 1.32 that affected loading of MPT models.
Repositories available
- 2, 3, 4, 5, 6 and 8-bit GGML models for CPU+GPU inference
- Unquantised fp16 model in pytorch format, for GPU inference and for further conversions
Prompt template
This is a non-fine-tuned base model. So it is designed for text completion, not following instructions, such as in the following example:
The meaning of life is
A note regarding context length: 8K
The base model has an 8K context length. It is not yet confirmed if the 8K context of this model works with the quantised files.
If it does, KoboldCpp supports 8K context if you manually set it to 8K by adjusting the text box above the slider:
It is currently unknown as to increased context is compatible with other MPT GGML clients.
If you have feedback on this, please let me know.
Compatibilty
These files are not compatible with text-generation-webui, llama.cpp, or llama-cpp-python.
Currently they can be used with:
- KoboldCpp, a powerful inference engine based on llama.cpp, with good UI and GPU accelerated support for MPT models: KoboldCpp
- The ctransformers Python library, which includes LangChain support: ctransformers
- The LoLLMS Web UI which uses ctransformers: LoLLMS Web UI
- rustformers' llm
- The example
mpt
binary provided with ggml
As other options become available I will endeavour to update them here (do let me know in the Community tab if I've missed something!)
Tutorial for using LoLLMS Web UI
Provided files
Name | Quant method | Bits | Size | Max RAM required | Use case |
---|---|---|---|---|---|
mpt-30b.ggmlv0.q4_0.bin | q4_0 | 4 | 16.85 GB | 19.35 GB | Original llama.cpp quant method, 4-bit. |
mpt-30b.ggmlv0.q4_1.bin | q4_1 | 4 | 18.73 GB | 21.23 GB | Original llama.cpp quant method, 4-bit. Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models. |
mpt-30b.ggmlv0.q5_0.bin | q5_0 | 5 | 20.60 GB | 23.10 GB | Original llama.cpp quant method, 5-bit. Higher accuracy, higher resource usage and slower inference. |
mpt-30b.ggmlv0.q5_1.bin | q5_1 | 5 | 22.47 GB | 24.97 GB | Original llama.cpp quant method, 5-bit. Even higher accuracy, resource usage and slower inference. |
mpt-30b.ggmlv0.q8_0.bin | q8_0 | 8 | 31.83 GB | 34.33 GB | Original llama.cpp quant method, 8-bit. Almost indistinguishable from float16. High resource use and slow. Not recommended for most users. |
Note: the above RAM figures assume no GPU offloading. If layers are offloaded to the GPU, this will reduce RAM usage and use VRAM instead.
Discord
For further support, and discussions on these models and AI in general, join us at:
Thanks, and how to contribute.
Thanks to the chirper.ai team!
I've had a lot of people ask if they can contribute. I enjoy providing models and helping people, and would love to be able to spend even more time doing it, as well as expanding into new projects like fine tuning/training.
If you're able and willing to contribute it will be most gratefully received and will help me to keep providing more models, and to start work on new AI projects.
Donaters will get priority support on any and all AI/LLM/model questions and requests, access to a private Discord room, plus other benefits.
- Patreon: https://patreon.com/TheBlokeAI
- Ko-Fi: https://ko-fi.com/TheBlokeAI
Special thanks to: Luke from CarbonQuill, Aemon Algiz, Dmitriy Samsonov.
Patreon special mentions: Mano Prime, Fen Risland, Derek Yates, Preetika Verma, webtim, Sean Connelly, Alps Aficionado, Karl Bernard, Junyu Yang, Nathan LeClaire, Chris McCloskey, Lone Striker, Asp the Wyvern, Eugene Pentland, Imad Khwaja, trip7s trip, WelcomeToTheClub, John Detwiler, Artur Olbinski, Khalefa Al-Ahmad, Trenton Dambrowitz, Talal Aujan, Kevin Schuppel, Luke Pendergrass, Pyrater, Joseph William Delisle, terasurfer , vamX, Gabriel Puliatti, David Flickinger, Jonathan Leane, Iucharbius , Luke, Deep Realms, Cory Kujawski, ya boyyy, Illia Dulskyi, senxiiz, Johann-Peter Hartmann, John Villwock, K, Ghost , Spiking Neurons AB, Nikolai Manek, Rainer Wilmers, Pierre Kircher, biorpg, Space Cruiser, Ai Maven, subjectnull, Willem Michiel, Ajan Kanaga, Kalila, chris gileta, Oscar Rangel.
Thank you to all my generous patrons and donaters!
Original model card: MosaicML's MPT-30B
MPT-30B
MPT-30B is a decoder-style transformer pretrained from scratch on 1T tokens of English text and code. This model was trained by MosaicML.
MPT-30B is part of the family of Mosaic Pretrained Transformer (MPT) models, which use a modified transformer architecture optimized for efficient training and inference.
MPT-30B comes with special features that differentiate it from other LLMs, including an 8k token context window (which can be further extended via finetuning; see MPT-7B-StoryWriter), support for context-length extrapolation via ALiBi, and efficient inference + training via FlashAttention. It also has strong coding abilities thanks to its pretraining mix. MPT models can also be served efficiently with both standard HuggingFace pipelines and NVIDIA's FasterTransformer. The size of MPT-30B was also specifically chosen to make it easy to deploy on a single GPU—either 1xA100-80GB in 16-bit precision or 1xA100-40GB in 8-bit precision.
This model uses the MosaicML LLM codebase, which can be found in the llm-foundry repository. It was trained by MosaicML’s NLP team on the MosaicML platform for LLM pretraining, finetuning, and inference.
How is this model different?
MPT-30B is:
- Licensed for the possibility of commercial use (unlike LLaMA).
- Trained on a large amount of data (1T tokens like LLaMA vs. 300B for Pythia, 300B for OpenLLaMA, and 800B for StableLM).
- Prepared to handle extremely long inputs thanks to ALiBi.
- Capable of fast training and inference (via FlashAttention and FasterTransformer)
- Equipped with highly efficient open-source training code via the llm-foundry repository
Models finetuned off MPT-30B:
The following models are finetuned on MPT-30B:
MPT-30B-Instruct: a model for short-form instruction following. Built by finetuning MPT-30B on several carefully curated datasets.
- License: CC-By-NC-SA-3.0
MPT-30B-Chat: a chatbot-like model for dialogue generation. Built by finetuning MPT-30B on ShareGPT-Vicuna, Camel-AI, GPTeacher, Guanaco, Baize and some generated datasets.
- License: CC-By-NC-SA-4.0
- Demo on Hugging Face Spaces
Model Date
June 22, 2023
Model License
Apache-2.0
Documentation
- Blog post: MPT-30B: Raising the bar for open-source foundation models
- Codebase (mosaicml/llm-foundry repo)
- Questions: Feel free to contact us via the MosaicML Community Slack!
How to Use
This model is best used with the MosaicML llm-foundry repository for training and finetuning.
import transformers
model = transformers.AutoModelForCausalLM.from_pretrained(
'mosaicml/mpt-30b',
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, ALiBi, QK LayerNorm, and more.
To use the optimized triton implementation of FlashAttention, you can load the model on GPU (cuda:0
) with attn_impl='triton'
and with bfloat16
precision:
import torch
import transformers
name = 'mosaicml/mpt-30b'
config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True)
config.attn_config['attn_impl'] = 'triton' # change this to use triton-based FlashAttention
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
)
The model was trained initially with a sequence length of 4096 with an additional pretraining stage for sequence length adapation up to 8192. However, ALiBi enables users to increase the maximum sequence length even further during finetuning and/or inference. For example:
import transformers
name = 'mosaicml/mpt-30b'
config = transformers.AutoConfig.from_pretrained(name, trust_remote_code=True)
config.max_seq_len = 16384 # (input + output) tokens can now be up to 16384
model = transformers.AutoModelForCausalLM.from_pretrained(
name,
config=config,
trust_remote_code=True
)
This model was trained with the MPT-30B tokenizer which is identical to the EleutherAI/gpt-neox-20b tokenizer.
from transformers import AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained('mosaicml/mpt-30b')
The model can then be used, for example, within a text-generation pipeline.
Note: when running Torch modules in lower precision, it is best practice to use the torch.autocast context manager.
from transformers import pipeline
with torch.autocast('cuda', dtype=torch.bfloat16):
inputs = tokenizer('Here is a recipe for vegan banana bread:\n', return_tensors="pt").to('cuda')
outputs = model.generate(**inputs, max_new_tokens=100)
print(tokenizer.batch_decode(outputs, skip_special_tokens=True))
# or using the HF pipeline
pipe = pipeline('text-generation', model=model, tokenizer=tokenizer, device='cuda:0')
with torch.autocast('cuda', dtype=torch.bfloat16):
print(
pipe('Here is a recipe for vegan banana bread:\n',
max_new_tokens=100,
do_sample=True,
use_cache=True))
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
- It uses ALiBi (Attention with Linear Biases) and does not use positional embeddings
- It does not use biases
Hyperparameter | Value |
---|---|
n_parameters | 29.95B |
n_layers | 48 |
n_heads | 64 |
d_model | 7168 |
vocab size | 50432 |
sequence length | 8192 |
Training Data
Streaming Datasets
Data was formatted using the MosaicML StreamingDataset 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 on the following data mix:
Data Source | Number of Tokens in Source | Proportion | Effective Number of Tokens | Epochs |
---|---|---|---|---|
mC4 3.1.0 - English (200+ words) | 2417.99 B | 33.50% | 335 B | 0.14 |
c4 - English - SemDedup 80% | 100.42 B | 29.90% | 299 B | 2.98 |
RedPajama - CommonCrawl | 878.45 B | 8.50% | 85 B | 0.097 |
The Stack - Selected Languages | 463.78 B | 10.00% | 100 B | 0.22 |
RedPajama - Wikipedia | 4.87 B | 4.00% | 40 B | 8.21 |
The Stack - Markdown | 107.07 B | 4.50% | 45 B | 0.42 |
Semantic Scholar ORC | 48.95 B | 3.30% | 33 B | 0.67 |
RedPajama - Books | 26.02 B | 3.00% | 30 B | 1.15 |
RedPajama - arXiv | 28.10 B | 1.90% | 19 B | 0.68 |
RedPajama - StackExchange | 20.54 B | 1.40% | 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 sequence length. To build 8k support into MPT-30B efficiently, we first pre-trained on 1T tokens using sequences that were 2k tokens long, and then trained for an additional 50B tokens using sequences that were 8k tokens long.
The data was tokenized using the 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).
Training Configuration
The model was trained in three stages using the MosaicML Platform: (i) First it was trained on 440 A100-40GBs with a batch size of 1760. (ii) Then, on 216 A100-40GBs with a batch size of 1728. (iii) Training was completed on 256 H100-80GBs with a batch size of 512 with 8k context length and 50B tokens. The model was trained with sharded data parallelism using FSDP and used the LION optimizer.
Limitations and Biases
The following language is modified from EleutherAI's GPT-NeoX-20B
MPT-30B (Base) is not intended for deployment without finetuning. It should not be used for human-facing interactions without further guardrails and user consent.
MPT-30B can produce factually incorrect output, and should not be relied on to produce factually accurate information. MPT-30B 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 and deploying your own MPT or LLMs on the MosaicML Platform, sign up here.
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 consult 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-30B: Raising the bar
for open-source foundation models},
year = {2023},
url = {www.mosaicml.com/blog/mpt-30b},
note = {Accessed: 2023-06-22},
urldate = {2023-06-22}
}
- Downloads last month
- 8