transformers/docs/source/en/model_doc/gpt_neo.md

GPT Neo

Overview

The GPTNeo model was released in the EleutherAI/gpt-neo repository by Sid Black, Stella Biderman, Leo Gao, Phil Wang and Connor Leahy. It is a GPT2 like causal language model trained on the Pile dataset.

The architecture is similar to GPT2 except that GPT Neo uses local attention in every other layer with a window size of 256 tokens.

This model was contributed by valhalla.

Usage example

The generate() method can be used to generate text using GPT Neo model.

>>> from transformers import GPTNeoForCausalLM, GPT2Tokenizer

>>> model = GPTNeoForCausalLM.from_pretrained("EleutherAI/gpt-neo-1.3B")
>>> tokenizer = GPT2Tokenizer.from_pretrained("EleutherAI/gpt-neo-1.3B")

>>> prompt = (
...     "In a shocking finding, scientists discovered a herd of unicorns living in a remote, "
...     "previously unexplored valley, in the Andes Mountains. Even more surprising to the "
...     "researchers was the fact that the unicorns spoke perfect English."
... )

>>> input_ids = tokenizer(prompt, return_tensors="pt").input_ids

>>> gen_tokens = model.generate(
...     input_ids,
...     do_sample=True,
...     temperature=0.9,
...     max_length=100,
... )
>>> gen_text = tokenizer.batch_decode(gen_tokens)[0]

Combining GPT-Neo and Flash Attention 2

First, make sure to install the latest version of Flash Attention 2 to include the sliding window attention feature.

pip install -U flash-attn --no-build-isolation

Make also sure that you have a hardware that is compatible with Flash-Attention 2. Read more about it in the official documentation of flash-attn repository. Make also sure to load your model in half-precision (e.g. `torch.float16``)

To load and run a model using Flash Attention 2, refer to the snippet below:

>>> import torch
>>> from transformers import AutoModelForCausalLM, AutoTokenizer
>>> device = "cuda" # the device to load the model onto

>>> model = AutoModelForCausalLM.from_pretrained("EleutherAI/gpt-neo-2.7B", torch_dtype=torch.float16, use_flash_attention_2=True)
>>> tokenizer = AutoTokenizer.from_pretrained("EleutherAI/gpt-neo-2.7B")

>>> prompt = "def hello_world():"

>>> model_inputs = tokenizer([prompt], return_tensors="pt").to(device)
>>> model.to(device)

>>> generated_ids = model.generate(**model_inputs, max_new_tokens=100, do_sample=True)
>>> tokenizer.batch_decode(generated_ids)[0]
"def hello_world():\n    >>> run_script("hello.py")\n    >>> exit(0)\n<|endoftext|>"

Expected speedups

Below is an expected speedup diagram that compares pure inference time between the native implementation in transformers using EleutherAI/gpt-neo-2.7B checkpoint and the Flash Attention 2 version of the model. Note that for GPT-Neo it is not possible to train / run on very long context as the max position embeddings is limited to 2048 - but this is applicable to all gpt-neo models and not specific to FA-2

Resources

• Text classification task guide
• Causal language modeling task guide

GPTNeoConfig

autodoc GPTNeoConfig

GPTNeoModel

autodoc GPTNeoModel - forward

GPTNeoForCausalLM

autodoc GPTNeoForCausalLM - forward

GPTNeoForQuestionAnswering

autodoc GPTNeoForQuestionAnswering - forward

GPTNeoForSequenceClassification

autodoc GPTNeoForSequenceClassification - forward

GPTNeoForTokenClassification

autodoc GPTNeoForTokenClassification - forward

FlaxGPTNeoModel

autodoc FlaxGPTNeoModel - call

FlaxGPTNeoForCausalLM

autodoc FlaxGPTNeoForCausalLM - call