Top news of the week: 15.08.2019.

#NeurIPS2019 #MineRL #NLP #WMT2019 #AI #ArtificialIntelligence #DeepLearning #MachineLearning #education #ML

On Aug 13, 2019
@demishassabis shared
RT @DeepMindAI: We are excited to release Behaviour Suite for Reinforcement Learning, or ‘bsuite’ – a collection of carefully-designed experiments that investigate core capabilities of RL agents GitHub: https://t.co/kGkRtFhezR Paper: https://t.co/UJcALddJpc
Open

Behaviour Suite for Reinforcement Learning (bsuite)

bsuite is a collection of carefully-designed experiments that investigate core capabilities of a reinforcement learning (RL) agent - deepmind/bsuite

On Aug 9, 2019
@erichorvitz shared
RT @MSFTResearch: Latest @Project_Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. See the role @CarnegieMellon students played in organizing the contest and sign up: https://t.co/xO5PGs3YfS #NeurIPS2019
Open

Microsoft Research Blog

Latest Project Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. Find out the role Carnegie ...

On Aug 13, 2019
@karpathy shared
RT @ctnzr: Here’s how we trained an 8.3B parameter GPT-2. We alternate row- and column- partitioning in the Transformer in order to remove synchronization and use hybrid model/data parallelism. 15 PFlops sustained on 512 GPUs. Details and code: https://t.co/7eXA6r15yX https://t.co/sEk4q0hU7T
Open

MegatronLM: Training Billion+ Parameter Language Models Using GPU Model Parallelism

We train an 8.3 billion parameter transformer language model with 8-way model parallelism and 64-way data parallelism on 512 GPUs, making it the largest transformer based language model ...

On Aug 14, 2019
@ylecun shared
A nice blog post explaining recent progress in NLP and translation and some Facebook AI's contributions. BERT, RoBERTa, backtranslation, GLUE, SuperGLUE, WMT explained.... https://t.co/sFl6VxdWx5
Open

New advances in natural language processing to better connect people

Recently, Facebook AI has advanced state-of-the-art results in key language understanding tasks and also launched a new benchmark to push AI systems further

On Aug 10, 2019
@weballergy shared
RT @alan_karthi: Brilliant piece from @drnigam and colleagues, on the need to consider downstream action/ clinical consequence from ML models in their evaluation. https://t.co/lzfcmgPewj
Open

Making Machine Learning Models Clinically Useful

This Viewpoint reviews conventional ways of assessing performance of machine learning models to diagnose or predict outcomes, but emphasizes that if machine learning is to improve patient ...

On Aug 13, 2019
@demishassabis shared
RT @DeepMindAI: We are excited to release Behaviour Suite for Reinforcement Learning, or ‘bsuite’ – a collection of carefully-designed experiments that investigate core capabilities of RL agents GitHub: https://t.co/kGkRtFhezR Paper: https://t.co/UJcALddJpc
Open

Behaviour Suite for Reinforcement Learning (bsuite)

bsuite is a collection of carefully-designed experiments that investigate core capabilities of a reinforcement learning (RL) agent - deepmind/bsuite

On Aug 13, 2019
@demishassabis shared
RT @DeepMindAI: We are excited to release Behaviour Suite for Reinforcement Learning, or ‘bsuite’ – a collection of carefully-designed experiments that investigate core capabilities of RL agents GitHub: https://t.co/kGkRtFhezR Paper: https://t.co/UJcALddJpc
Open

Behaviour Suite for Reinforcement Learning (bsuite)

bsuite is a collection of carefully-designed experiments that investigate core capabilities of a reinforcement learning (RL) agent - deepmind/bsuite

On Aug 9, 2019
@erichorvitz shared
RT @MSFTResearch: Latest @Project_Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. See the role @CarnegieMellon students played in organizing the contest and sign up: https://t.co/xO5PGs3YfS #NeurIPS2019
Open

Microsoft Research Blog

Latest Project Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. Find out the role Carnegie ...

On Aug 13, 2019
@karpathy shared
RT @ctnzr: Here’s how we trained an 8.3B parameter GPT-2. We alternate row- and column- partitioning in the Transformer in order to remove synchronization and use hybrid model/data parallelism. 15 PFlops sustained on 512 GPUs. Details and code: https://t.co/7eXA6r15yX https://t.co/sEk4q0hU7T
Open

MegatronLM: Training Billion+ Parameter Language Models Using GPU Model Parallelism

We train an 8.3 billion parameter transformer language model with 8-way model parallelism and 64-way data parallelism on 512 GPUs, making it the largest transformer based language model ...

On Aug 14, 2019
@ylecun shared
A nice blog post explaining recent progress in NLP and translation and some Facebook AI's contributions. BERT, RoBERTa, backtranslation, GLUE, SuperGLUE, WMT explained.... https://t.co/sFl6VxdWx5
Open

New advances in natural language processing to better connect people

Recently, Facebook AI has advanced state-of-the-art results in key language understanding tasks and also launched a new benchmark to push AI systems further

On Aug 10, 2019
@weballergy shared
RT @alan_karthi: Brilliant piece from @drnigam and colleagues, on the need to consider downstream action/ clinical consequence from ML models in their evaluation. https://t.co/lzfcmgPewj
Open

Making Machine Learning Models Clinically Useful

This Viewpoint reviews conventional ways of assessing performance of machine learning models to diagnose or predict outcomes, but emphasizes that if machine learning is to improve patient ...

On Aug 13, 2019
@demishassabis shared
RT @DeepMindAI: We are excited to release Behaviour Suite for Reinforcement Learning, or ‘bsuite’ – a collection of carefully-designed experiments that investigate core capabilities of RL agents GitHub: https://t.co/kGkRtFhezR Paper: https://t.co/UJcALddJpc
Open

Behaviour Suite for Reinforcement Learning (bsuite)

bsuite is a collection of carefully-designed experiments that investigate core capabilities of a reinforcement learning (RL) agent - deepmind/bsuite

On Aug 9, 2019
@erichorvitz shared
RT @MSFTResearch: Latest @Project_Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. See the role @CarnegieMellon students played in organizing the contest and sign up: https://t.co/xO5PGs3YfS #NeurIPS2019
Open

Microsoft Research Blog

Latest Project Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. Find out the role Carnegie ...

On Aug 13, 2019
@karpathy shared
RT @ctnzr: Here’s how we trained an 8.3B parameter GPT-2. We alternate row- and column- partitioning in the Transformer in order to remove synchronization and use hybrid model/data parallelism. 15 PFlops sustained on 512 GPUs. Details and code: https://t.co/7eXA6r15yX https://t.co/sEk4q0hU7T
Open

MegatronLM: Training Billion+ Parameter Language Models Using GPU Model Parallelism

We train an 8.3 billion parameter transformer language model with 8-way model parallelism and 64-way data parallelism on 512 GPUs, making it the largest transformer based language model ...

On Aug 14, 2019
@ylecun shared
A nice blog post explaining recent progress in NLP and translation and some Facebook AI's contributions. BERT, RoBERTa, backtranslation, GLUE, SuperGLUE, WMT explained.... https://t.co/sFl6VxdWx5
Open

New advances in natural language processing to better connect people

Recently, Facebook AI has advanced state-of-the-art results in key language understanding tasks and also launched a new benchmark to push AI systems further

On Aug 10, 2019
@weballergy shared
RT @alan_karthi: Brilliant piece from @drnigam and colleagues, on the need to consider downstream action/ clinical consequence from ML models in their evaluation. https://t.co/lzfcmgPewj
Open

Making Machine Learning Models Clinically Useful

This Viewpoint reviews conventional ways of assessing performance of machine learning models to diagnose or predict outcomes, but emphasizes that if machine learning is to improve patient ...

On Aug 13, 2019
@demishassabis shared
RT @DeepMindAI: We are excited to release Behaviour Suite for Reinforcement Learning, or ‘bsuite’ – a collection of carefully-designed experiments that investigate core capabilities of RL agents GitHub: https://t.co/kGkRtFhezR Paper: https://t.co/UJcALddJpc
Open

Behaviour Suite for Reinforcement Learning (bsuite)

bsuite is a collection of carefully-designed experiments that investigate core capabilities of a reinforcement learning (RL) agent - deepmind/bsuite

On Aug 9, 2019
@erichorvitz shared
RT @MSFTResearch: Latest @Project_Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. See the role @CarnegieMellon students played in organizing the contest and sign up: https://t.co/xO5PGs3YfS #NeurIPS2019
Open

Microsoft Research Blog

Latest Project Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. Find out the role Carnegie ...

On Aug 13, 2019
@karpathy shared
RT @ctnzr: Here’s how we trained an 8.3B parameter GPT-2. We alternate row- and column- partitioning in the Transformer in order to remove synchronization and use hybrid model/data parallelism. 15 PFlops sustained on 512 GPUs. Details and code: https://t.co/7eXA6r15yX https://t.co/sEk4q0hU7T
Open

MegatronLM: Training Billion+ Parameter Language Models Using GPU Model Parallelism

We train an 8.3 billion parameter transformer language model with 8-way model parallelism and 64-way data parallelism on 512 GPUs, making it the largest transformer based language model ...

On Aug 14, 2019
@ylecun shared
A nice blog post explaining recent progress in NLP and translation and some Facebook AI's contributions. BERT, RoBERTa, backtranslation, GLUE, SuperGLUE, WMT explained.... https://t.co/sFl6VxdWx5
Open

New advances in natural language processing to better connect people

Recently, Facebook AI has advanced state-of-the-art results in key language understanding tasks and also launched a new benchmark to push AI systems further

On Aug 10, 2019
@weballergy shared
RT @alan_karthi: Brilliant piece from @drnigam and colleagues, on the need to consider downstream action/ clinical consequence from ML models in their evaluation. https://t.co/lzfcmgPewj
Open

Making Machine Learning Models Clinically Useful

This Viewpoint reviews conventional ways of assessing performance of machine learning models to diagnose or predict outcomes, but emphasizes that if machine learning is to improve patient ...

On Aug 9, 2019
@erichorvitz shared
RT @MSFTResearch: Latest @Project_Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. See the role @CarnegieMellon students played in organizing the contest and sign up: https://t.co/xO5PGs3YfS #NeurIPS2019
Open

Microsoft Research Blog

Latest Project Malmo-based competition challenges participants to create agents capable of mining a diamond in Minecraft with limited training time and data. Find out the role Carnegie ...

On Aug 13, 2019
@karpathy shared
RT @ctnzr: Here’s how we trained an 8.3B parameter GPT-2. We alternate row- and column- partitioning in the Transformer in order to remove synchronization and use hybrid model/data parallelism. 15 PFlops sustained on 512 GPUs. Details and code: https://t.co/7eXA6r15yX https://t.co/sEk4q0hU7T
Open

MegatronLM: Training Billion+ Parameter Language Models Using GPU Model Parallelism

We train an 8.3 billion parameter transformer language model with 8-way model parallelism and 64-way data parallelism on 512 GPUs, making it the largest transformer based language model ...

On Aug 14, 2019
@ylecun shared
A nice blog post explaining recent progress in NLP and translation and some Facebook AI's contributions. BERT, RoBERTa, backtranslation, GLUE, SuperGLUE, WMT explained.... https://t.co/sFl6VxdWx5
Open

New advances in natural language processing to better connect people

Recently, Facebook AI has advanced state-of-the-art results in key language understanding tasks and also launched a new benchmark to push AI systems further

On Aug 10, 2019
@weballergy shared
RT @alan_karthi: Brilliant piece from @drnigam and colleagues, on the need to consider downstream action/ clinical consequence from ML models in their evaluation. https://t.co/lzfcmgPewj
Open

Making Machine Learning Models Clinically Useful

This Viewpoint reviews conventional ways of assessing performance of machine learning models to diagnose or predict outcomes, but emphasizes that if machine learning is to improve patient ...

On Aug 8, 2019
@jeremyphoward shared
RT @bendee983: Read my latest explainer on @PCMag: What is deep learning? (includes great thoughts from @jeremyphoward on what makes deep learning fascinating) https://t.co/nl4CjZYKqh #ArtificialIntelligence #DeepLearning #MachineLearning
Open

What Is Deep Learning?

Deep learning is a subset of machine learning, a branch of artificial intelligence that configures computers to perform tasks through experience. Here's a deep dive.

On Aug 10, 2019
@pabbeel shared
Was fun chatting with @byronreese on Voices in AI – Episode 93: A Conversation with Pieter Abbeel https://t.co/2wnICPqtnO via @Gigaom
Open

Voices in AI – Episode 93: A Conversation with Pieter Abbeel

About this Episode Episode 93 of Voices in AI features Byron speaking with Berkeley Robotic Learning Lab Director Pieter Abbeel about…

On Aug 13, 2019
@pabbeel shared
RT @DisruptorDaily: Arjun Singh let us in on the ways that @gradescope is changing #education with #AI Read More: https://t.co/DltPZREeEN
Open

AI In Education Use Case #7: Gradescope

  This interview is part of our new AI in Education series, where… ...

On Aug 12, 2019
@_rockt shared
RT @WonderMicky: Now thru Oct 4, students in any year of their PhD who are engaged in innovative and relevant research in areas related to CS and engineering at an accredited university (anywhere in the world!) are encouraged to apply to the @facebook fellowship program! https://t.co/8BI6wdXY8s
Open

Fellowship Program on Facebook

The Facebook Fellowship Program is designed to encourage and support promising doctoral students who are engaged in innovative and relevant research in areas related to computer science and ...

On Aug 13, 2019
@thinkmariya shared
Matthew Opala explained in the details the obstacles faced in a machine learning projects. #AI #ML https://t.co/kxUcq7EO5k
Open

Major Challenges for Machine Learning Projects

Understand the limits of contemporary machine learning technology Many companies face the challenge of educating customers on the possible applications of their innovative technology. A ...