Mpt 30b

🚀 MPT-30B

MPT-30B is a decoder-style transformer that has been pretrained from scratch on 1T tokens of English text and code. It was developed by MosaicML. This model is part of the Mosaic Pretrained Transformer (MPT) family, which features a modified transformer architecture optimized for efficient training and inference.

🚀 Quick Start

MPT-30B is best used with the MosaicML llm-foundry repository for training and finetuning.

Basic Usage

import transformers
model = transformers.AutoModelForCausalLM.from_pretrained(
  'mosaicml/mpt-30b',
  trust_remote_code=True
)

Advanced Usage

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 2048 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))

✨ Features

• Commercial Use License: Licensed for the possibility of commercial use (unlike LLaMA).
• Large-Scale Training Data: Trained on a large amount of data (1T tokens like LLaMA vs. 300B for Pythia, 300B for OpenLLaMA, and 800B for StableLM).
• Long Input Handling: Prepared to handle extremely long inputs thanks to ALiBi.
• Efficient Training and Inference: Capable of fast training and inference (via FlashAttention and FasterTransformer).
• Efficient Open-Source Training Code: Equipped with highly efficient open-source training code via the llm-foundry repository.
• 8k Token Context Window: Comes with an 8k token context window (which can be further extended via finetuning; see MPT-7B-StoryWriter).
• Context-Length Extrapolation: Supports context-length extrapolation via ALiBi.
• Single GPU Deployment: The size of MPT-30B was 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.

📚 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!

🔧 Technical Details

Model Architecture

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.

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:

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.

📄 License

The model is licensed under Apache-2.0.

💡 Usage Tip

⚠️ Important 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.

💡 Usage Tip

When running Torch modules in lower precision, it is best practice to use the torch.autocast context manager.

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}
}