Motion prediction with Lyft

https://self-driving.lyft.com/level5/prediction/

Dataset (A collection of traffic agent motion data including the logs of movement of cars, cyclists, pedestrians, and other traffic agents encountered by Lyft's autonomous fleet:
https://self-driving.lyft.com/level5/prediction/

The testing and validation sets will be released as part of the Lyft competition, which will kick off in August on the Kaggle platform.

Blog post:
https://medium.com/lyftlevel5/fueling-self-driving-research-with-level-5s-open-prediction-dataset-f0175e2b0cf8

ICRA 2020 Keynote - Can Deep Reinforcement Learning from pixels be made as efficient as from state?


https://ieeetv.ieee.org/mobile/video/icra-2020-keynote-can-deep-reinforcement-learning-from-pixels-be-made-as-efficient-as-from-state#.Xv7cCLWTGHQ.telegram

Here is the RL tutorial at EEMLcommunity summer school to teach value based methods. Tabular agents are in python & function approximation agents are in JAX using Acme (best viewed in Colab).

https://colab.research.google.com/github/eemlcommunity/PracticalSessions2020/blob/master/rl/EEML2020_RL_Tutorial.ipynb

Duality — A New Approach to Reinforcement Learning

https://ai.googleblog.com/2020/07/duality-new-approach-to-reinforcement.html?m=1

Deep Reinforcement Learning and
its Neuroscientific Implications

https://syncedreview.com/2020/07/10/deepmind-explores-deep-rl-for-brain-and-behaviour-research/

Paper:
https://arxiv.org/pdf/2007.03750.pdf

Scalable Deep Generative Modeling for Sparse Graphs

Paper:
https://arxiv.org/abs/2006.15502

Code:
https://github.com/google-research/google-research/tree/master/bigg

gpt-3
یه توضیح برای layout می‌گیره و کد jsx براش تولید می‌کنه!

https://twitter.com/sharifshameem/status/1282676454690451457?s=20

#twitter

منبع
@nlp_stuff

استفاده از مدل GPT-2 برای تولید متن بلاگ‌-پست فارسی توسط آقای افشین خاشعی از مدیران ارشد توسعه نرم افزار در Amazon:

مثال اول:
https://lnkd.in/gxvxvuT
مثال دوم:
https://lnkd.in/gxniKcT
مثال سوم:
https://lnkd.in/gyyJcTk

توضیحات فنی:
https://medium.com/@khashei/a-not-so-dangerous-ai-in-the-persian-language-39172a641c84

همونطور که میدونید اخیرا انتشار مدل زبانی GPT-3 سر و صدای زیادی به پا کرده. توی این ویدئو ازش برای یه کاربرد جالب مثل نوشتن کوئری استفاده شده...

https://twitter.com/FaraazNishtar/status/1285934622891667457?s=20

میتونید درباره این مدل بیشتر بخونید:
https://openai.com/blog/openai-api/

ALBERT-Persian: A Lite BERT for Self-supervised Learning of Language Representations for the Persian Language

Try:
https://albert-lab.m3hrdadfi.me/

GitHub:
https://github.com/m3hrdadfi/albert-persian#albert-persian-a-lite-bert-for-self-supervised-learning-of-language-representations-for-the-persian-language

GPU and TPU-accelerated NumPy API, interoperable with the rest of the TF ecosystem.
https://www.tensorflow.org/api_docs/python/tf/experimental/numpy

Face detection with TFLite in Flutter:

https://medium.com/@mundorap2010/face-detection-with-tflite-model-without-firebase-in-flutter-6eadf888f3b0

handcalcs:
Python calculations in Jupyter,
as though you wrote them by hand.

$ 𝚙𝚒𝚙 𝚒𝚗𝚜𝚝𝚊𝚕𝚕 𝚑𝚊𝚗𝚍𝚌𝚊𝚕𝚌𝚜

GitHub:
https://github.com/connorferster/handcalcs

هوشمندی فارغ از اینکه مصنوعی باشه یا عمومی، واقعا چیه و چطوری اندازه‌گیری میشه؟ از فرانسوا شوله، نویسنده کراس و از مهندسین گوگل، توی این پادکست بشنوید:


https://www.youtube.com/watch?v=PUAdj3w3wO4