# forck.live — 27 June – 3 July 2022 > AI lab announcements, news reports and podcasts. forck.live follows what companies announce about their own products. Press may corroborate an official story; it never leads and never appears as an item of its own. A missing date stays missing. A title is the publisher's words — where a company does not publish in English, an English rendering is listed separately as `Title in English` and named as this site's translation, never as theirs. This file is one issue of the weekly archive: every confirmed first-party announcement of ISO week 2022-W26, 27 June – 3 July 2022, newest first. The window never moves, so this URL always names the same seven days — but a post published inside them and detected later is added when it arrives, so a closed week can still gain entries. Every entry is one first-party announcement, and carries the same eight fields in this order before any others: Company · Announced · Category · Coverage · Announcement · Group · Source · Record `Source` is the company's own page the announcement was traced to and `Record` is this site's page for it. Both are safe to cite. `Coverage` is a number of outlets, or `not counted` — which means this record reached no count of its own, never that nobody wrote about it. `Category` is this tracker's own analysis, and is authoritative wherever it is stated. `not stated` means either that nothing has analysed the post yet or that the analysis placed it outside this taxonomy; neither is a verdict about the announcement. `Announcement` is whether the post is news of something the company did, rather than a how-to, a customer story, a changelog listing or a staffing note. `Group` is which of this record's groups the entry falls in, carried on the entry itself so it holds where the groups are headings and where they are not. One of: `models`, `covered`, `announcements`, `routine`. Anything after `Record` is held detail that only some announcements carry. - This week's page: https://forck.live/week/2022-W26 - The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md - Catalogue: https://forck.live/llms.txt ## Announcements ### Hugging Face — Policy Gradient with PyTorch - Company: Hugging Face (huggingface.co) - Announced: 2022-06-30 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/deep-rl-pg - Record: https://forck.live/items/2181-policy-gradient-with-pytorch - Subject: Platform This is an educational article about Policy Gradient methods in reinforcement learning, specifically the Reinforce algorithm, with implementation using PyTorch. It is part of a free Deep Reinforcement Learning course by Hugging Face. ### Hugging Face — Liftoff! How to get started with your first ML project 🚀 - Company: Hugging Face (huggingface.co) - Announced: 2022-06-29 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/your-first-ml-project - Record: https://forck.live/items/2182-liftoff-how-to-get-started-with-your-first-ml-project - Subject: Platform Tips for beginners on starting their first machine learning project, using the Sentence Transformers library as an example. It covers what Sentence Transformers is, why to learn it, and a strategy for tackling a first project. ### OpenAI — DALL·E 2 pre-training mitigations - Company: OpenAI (openai.com) - Announced: 2022-06-28T07:00:00+00:00 - Category: safety-policy-update - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/dall-e-2-pre-training-mitigations - Record: https://forck.live/items/901-dall-e-2-pre-training-mitigations - Subject: GPT / ChatGPT / API OpenAI describes pre-training mitigations for DALL·E 2, including guardrails to prevent generated images from violating content policy. ### Hugging Face — Announcing Evaluation on the Hub - Company: Hugging Face (huggingface.co) - Announced: 2022-06-28 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/eval-on-the-hub - Record: https://forck.live/items/2184-announcing-evaluation-on-the-hub - Subject: Platform Hugging Face announced Evaluation on the Hub, a new tool that allows users to evaluate any model on any dataset on the Hub without writing code, powered by AutoTrain. ### Hugging Face — Accelerate Large Model Training using DeepSpeed - Company: Hugging Face (huggingface.co) - Announced: 2022-06-28 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/accelerate-deepspeed - Record: https://forck.live/items/2183-accelerate-large-model-training-using-deepspeed - Subject: Platform To use the Accelerate library with DeepSpeed ZeRO to train large models, enabling larger batch sizes and faster training without code changes. ### LG AI Research — LG AI Expert Course to Strengthen AI Capability Through Learning and Collaboration - Company: LG AI Research (lgresearch.ai) - Announced: 2022-06-27 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.lgresearch.ai/news/view?seq=220 - Record: https://forck.live/items/11069-lg-ai-expert-course-to-strengthen-ai-capability-through-learning-and - Subject: EXAONE LG AI Research held a completion ceremony for the first half of 2022's LG AI Expert Course, a 12-week AI training program for LG Group employees. Fifteen trainees from six affiliates completed the program, which includes one-to-one mentoring from LG AI Research specialists. Three research projects were awarded Best Practice honors, including a battery defect detection method, a refrigerator abnormality classification model, and an AI model for drug discovery. ### LG AI Research — LG AI Expert Course to Strengthen AI Capability Through Learning and Collaboration - Company: LG AI Research (lgresearch.ai) - Announced: 2022-06-27 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://www.lgresearch.ai/blog/view?seq=220 - Record: https://forck.live/items/4947-lg-ai-expert-course-to-strengthen-ai-capability-through-learning-and - Subject: EXAONE LG AI Research held a completion ceremony for its 2022 LG AI Expert Course, a 12-week training program for employees to develop AI skills through mentoring and practical projects, with awards for best research projects.