AWS Machine Learning Blog

MAY 8, 2023

With kernel auto-tuning, the engine selects the best algorithm for the target GPU, maximizing hardware utilization. SageMaker MMEs can horizontally scale using an auto scaling policy and provision additional GPU compute instances based on specified metrics. Note that the cell takes around 30 minutes to complete. !docker

ML 83

ML BERT Deep Learning Auto-complete 83

Explosion

JULY 6, 2021

SpanCategorizer for predicting arbitrary and overlapping spans A common task in applied NLP is extracting spans of texts from documents, including longer phrases or nested expressions. adds 5 new pipeline packages, including a new core family for Catalan and a new transformer-based pipeline for Danish using the danish-bert-botxo weights.

BERT 52

BERT Auto-complete NLP Python 52

Google Research AI blog

FEBRUARY 28, 2023

Dataset Description Auto-Arborist A multiview urban tree classification dataset that consists of ~2.6M MultiBERTs Predictions on Winogender Predictions of BERT on Winogender before and after several different interventions. UGIF A multi-lingual, multi-modal UI grounded dataset for step-by-step task completion on the smartphone.

Robotics 72

Robotics Deep Learning Auto-classification ML 72

Explosion

JANUARY 31, 2021

It’s much easier to configure and train your pipeline, and there are lots of new and improved integrations with the rest of the NLP ecosystem. And since modern NLP workflows often consist of multiple steps, there’s a new workflow system to help you keep your work organized. See NLP-progress for more results. Flair 2 89.7

NLP 52

NLP Python BERT Auto-classification 52

Jay Alammar

DECEMBER 16, 2020

This article focuses on auto-regressive models, but these methods are applicable to other architectures and tasks as well. Multiple methods exist for assigning importance scores to the inputs of an NLP model. A breakdown of this architecture is provided here. This is the first article in the series. The model used here is DistilGPT2.

Explainability 52

Explainability Auto-complete Auto-classification Neural Network 52

AWS Machine Learning Blog

APRIL 17, 2023

SageMaker LMI containers includes model download optimization by using the s5cmd library to speed up the model download time and container startup times, and eventually speed up auto scaling on SageMaker. A complete example that illustrates the no-code option can be found in the following notebook.

Prompt Engineer 78

Prompt Engineer Prompt Engineering Deep Learning Large Language Models 78

Snorkel AI

MARCH 2, 2023

It came to its own with the creation of the transformer architecture: Google’s BERT, OpenAI, GPT2 and then 3, LaMDA for conversation, Mina and Sparrow from Google DeepMind. As we look at the progression, we see that these state-of-the-art NLP models are getting larger and larger over time. So there’s obviously an evolution.

Large Language Models 52

Large Language Models Prompt Engineer Prompt Engineering Neural Network 52

Snorkel AI

MARCH 2, 2023

It came to its own with the creation of the transformer architecture: Google’s BERT, OpenAI, GPT2 and then 3, LaMDA for conversation, Mina and Sparrow from Google DeepMind. As we look at the progression, we see that these state-of-the-art NLP models are getting larger and larger over time. So there’s obviously an evolution.

Large Language Models 52

Large Language Models Prompt Engineer Prompt Engineering Neural Network 52

AWS Machine Learning Blog

MARCH 6, 2023

Large language models (LLMs) are neural network-based language models with hundreds of millions ( BERT ) to over a trillion parameters ( MiCS ), and whose size makes single-GPU training impractical. The preparation of a natural language processing (NLP) dataset abounds with share-nothing parallelism opportunities.

Large Language Models 78

Large Language Models LLM Machine Learning ML 78

AWS Machine Learning Blog

FEBRUARY 24, 2023

These include computer vision (CV), natural language processing (NLP), and generative AI models. Taking NLP models as an example, many of them exceed billions of parameters, which requires GPUs to satisfy low latency and high throughput requirements. We tested two NLP models: bert-base-uncased (109M) and roberta-large (335M).

BERT 87

BERT NLP Computer Vision Neural Network 87

AWS Machine Learning Blog

NOVEMBER 1, 2023

Additionally, you benefit from advanced features like auto scaling of inference endpoints, enhanced security, and built-in model monitoring. The pre-training of IDEFICS-9b took 350 hours to complete on 128 Nvidia A100 GPUs, whereas fine-tuning of IDEFICS-9b-instruct took 70 hours on 128 Nvidia A100 GPUs, both on AWS p4.24xlarge instances.

Auto-classification 92

Auto-classification LLM Auto-complete Generative AI 92

AWS Machine Learning Blog

JANUARY 9, 2023

The models can be completely heterogenous, with their own independent serving stack. After these business logic steps are complete, the inputs are passed through to ML models. It can especially be handy in cases with NLP and computer vision, where there are large payloads that require longer preprocessing times.

ML 74

ML Auto-complete Auto-classification Deep Learning 74