robocrunch
Philipp Schmid
@_philschmid
Machine Learning Engineer & Tech Lead at @huggingface 👨🏻💻 🤗 | Cloud enthusiast | 📍Nuremberg | 🇩🇪 https://t.co/l1ppq3q3hk
Tweets by Philipp Schmid
Optimum v1.2 adds ACCELERATED inference pipelines - including text generation - for @onnxruntime🚀 Learn how to accelerate RoBERTa for Question-Answering including quantization and optimization with 🤗Optimum in our blog 🦾🔥 📕https://t.co/CMSdeV0bLd ⭐️https://t.co/GwAHZMWa0a
Shared by
Philipp Schmid
at
5/10/2022
Did you know you could use Wav2vec2 for serverless automatic speech recognition (speech-to-text) with Hugging Face Transformers and SageMaker Serverless? 🤯🚀🗣 Check out the code snippet below. 👉🏻 https://t.co/8p3sELj0Jf
Shared by
Philipp Schmid
at
5/6/2022
Transformers are changing machine learning, starting with NLP, and now, with audio and computer vision💬👄 👀 You can now use the Hugging Face Inference DLC to do automatic speech recognition using wav2vec2 model or WavLM🤯 🖼 https://t.co/WSOwzFXwPv 📈 https://t.co/wpoDz3hT9j
Shared by
Philipp Schmid
at
4/28/2022
When to use SageMaker Serverless 🤔🤔 We ran different inference experiments transformers, like DistilBERT, Pegasus, and BART to measure their latency and cost.🧪⏱ Spoiler: 1 million requests with DistilBERT cost around 22$ 🤯 https://t.co/TxZaogst3S
Shared by
Philipp Schmid
at
4/26/2022
General Availability of SageMaker Serverless 🚀 We have worked very closely with AWS to ensure Hugging Face Transformers work out-of-the-box and are easy to deploy. 🤝 🤗 🖼 https://t.co/TdrbiUYr0y 📈 https://t.co/ckbUauGnHx
Shared by
Philipp Schmid
at
4/21/2022
Are you interested in accelerated highly optimized semantic search with Hugging Face and AWS🧐 We created a new e2e example on how to use Sentence Transformers with Hugging Face and AWS Inferentia to achieve < 5ms latency 🏎 🖼https://t.co/I6UElb0BL8 📈https://t.co/a5KdLVg22Y
Shared by
Philipp Schmid
at
4/19/2022
📣 📣 Last Call 📣 📣 Tomorrow is our hands-on workshop on how to "Accelerate BERT Inference with Knowledge Distillation & AWS Inferentia". 🏎 👉https://t.co/OGHTicgZMU
Shared by
Philipp Schmid
at
4/12/2022
📅 April 13, 2022 | 9:00AM PDT | 12:00 EST | 6:00PM CEST! Free 1-hour workshop on "Accelerate BERT Inference with Knowledge Distillation & AWS Inferentia".🏎 Join @_lewtun, @HeikoHotz, and me next Wednesday!🙋🏻♂️ 👉 https://t.co/OGHTicgZMU htt
Shared by
Philipp Schmid
at
4/5/2022
Saving cost is always interesting, especially when using GPUs.💾💰 By configuring 3 additional parameters you can save up to 90% training cost with Amazon SageMaker and Hugging Face Transformers. 😍 🖼 https://t.co/hyM0pJJ68x 📓 https://t.co/iJGe2mF32g
Shared by
Philipp Schmid
at
3/22/2022
Interested in getting the best cost-performance on AWS for your Hugging Face Transformers⁉️ Check out our new e2e tutorial on how to speed up BERT inference for text classification with Hugging Face and AWS🐌 ➡️ 🏎 🖼blog: https://t.co/AA5fU8HepV 📈nb: https://t.co/thoIyTw5zw h
Shared by
Philipp Schmid
at
3/16/2022
New Hugging Face SageMaker Example💫✨ Learn how to use a custom inference script to easily compute sentence embeddings using https://t.co/CJKx4IpSkb Sentence Transformer models. 🧮 🖼 blog: https://t.co/VebQxHH03D 📈 notebook: https://t.co/ef7dJBczI7
Shared by
Philipp Schmid
at
3/8/2022
New Hugging Face Amazon SageMaker Terraform Module update 💫✨ - 🆕 support for autoscaling your endpoints - 🆕 support for Asynchronous Inference Endpoints - 🆕 support for private Hugging Face Transformers from https://t.co/OHE3JJsMU6 - 🆕 improves naming with a random suffix
Shared by
Philipp Schmid
at
3/1/2022
⚠️New Example Alert⚠️ Learn how to deploy multiple HuggingFace Transformers for inference with Amazon SageMaker and Multi-Container Endpoints🔢 This can help you to improve endpoint utilization as well as optimize costs💾💰 👉🏻https://t.co/sv6xHkfhoY 👉🏻 https://t.co/o2TmQtxxq9 htt
Shared by
Philipp Schmid
at
2/22/2022
Asynchronous Inference Endpoints queues incoming requests and processes them asynchronously🧵 Check out this example on async inference with HuggingFace Transformers and learn how to: - deploy🚀 - request📩 - scale to zero📉 👉🏻 https://t.co/txPXrlRYNZ 👉🏻 https://t.co/DlAZL6R6TC h
Shared by
Philipp Schmid
at
2/15/2022
Summarization is the task of producing a shorter version of a document while preserving its important information🧠 Check out this new example using Keras, Tensorflow and Hugging Face Transformers, including scaling the training with Amazon SageMaker📈 👉🏻https://t.co/v9wiBb1nzH
Shared by
Philipp Schmid
at
1/19/2022
Last week the second part of the Hugging Face Course has been released🚀 👉🏻 https://t.co/KynaGomGkI We also reworked the "TensorFlow Getting Started with Amazon SageMaker" to use the new datasets utils to_tf_dataset to convert your dataset to TF🧑🏻💻👩🏼💻 👉🏻 https://t.co/i7e1DuHlux https:/
Shared by
Philipp Schmid
at
11/23/2021