# forck.live — 12–18 June 2023 > 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 2023-W24, 12–18 June 2023, 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/2023-W24 - The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md - Catalogue: https://forck.live/llms.txt ## Announcements ### Hugging Face — Yes, Transformers are Effective for Time Series Forecasting (+ Autoformer) - Company: Hugging Face (huggingface.co) - Announced: 2023-06-16 - Category: developer-tool-release - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/autoformer - Record: https://forck.live/items/2039-yes-transformers-are-effective-for-time-series-forecasting-autoformer - Subject: Platform The blog post argues that Transformers are effective for time series forecasting, introduces the Autoformer model, and announces that Autoformer is now available in the 🤗 Transformers library. It also presents benchmark results showing Autoformer outperforming DLinear. ### Hugging Face — Announcing our new Content Guidelines and Policy - Company: Hugging Face (huggingface.co) - Announced: 2023-06-15 - Category: safety-policy-update - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/content-guidelines-update - Record: https://forck.live/items/2042-announcing-our-new-content-guidelines-and-policy - Subject: Platform Hugging Face announced updated Community Policy and Content Guidelines, focusing on consent, respect for rights, and maintaining a welcoming environment for the community. ### Hugging Face — Deploy Livebook notebooks as apps to Hugging Face Spaces - Company: Hugging Face (huggingface.co) - Announced: 2023-06-15 - Category: capability-change - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/livebook-app-deployment - Record: https://forck.live/items/2041-deploy-livebook-notebooks-as-apps-to-hugging-face-spaces - Subject: Platform Hugging Face announces that Livebook notebooks can now be deployed as apps to Hugging Face Spaces, enabling Elixir-based machine learning applications with features like integration with Hugging Face Models, multiplayer ML apps, and concurrent model serving. ### Hugging Face — Faster Stable Diffusion with Core ML on iPhone, iPad, and Mac - Company: Hugging Face (huggingface.co) - Announced: 2023-06-15 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/fast-diffusers-coreml - Record: https://forck.live/items/2040-faster-stable-diffusion-with-core-ml-on-iphone-ipad-and-mac - Subject: Platform Apple announced Core ML optimizations at WWDC'23, including 6-bit palettization, making Stable Diffusion run faster and with less memory on iPhone, iPad, and Mac. Hugging Face has converted and uploaded the four official Stable Diffusion models in the optimized Core ML format. ### OpenAI — Function calling and other API updates - Company: OpenAI (openai.com) - Announced: 2023-06-13T07:00:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/function-calling-and-other-api-updates - Record: https://forck.live/items/856-function-calling-and-other-api-updates - Subject: GPT / ChatGPT / API OpenAI announces API updates including more steerable models, function calling, longer context, and lower prices. ### Hugging Face — Hugging Face and AMD partner on accelerating state-of-the-art models for CPU and GPU platforms - Company: Hugging Face (huggingface.co) - Announced: 2023-06-13 - Category: partnership-acquisition - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/huggingface-and-amd - Record: https://forck.live/items/2043-hugging-face-and-amd-partner-on-accelerating-state-of-the-art-models-for-cpu - Subject: Platform Hugging Face announces that AMD has joined its Hardware Partner Program to collaborate on optimizing transformer models for AMD CPUs, GPUs (Instinct MI2xx, MI3xx, Radeon Navi3x), and the Alveo V70 AI accelerator, with plans to support architectures such as BERT, GPT-2, and LLaMA, and frameworks including PyTorch and TensorFlow. ### Replit — June 6, 2023, Single-Sign-On Security Vulnerability - Company: Replit (replit.com) - Announced: 2023-06-12T12:00:00+00:00 - Category: safety-policy-update - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://replit.com/blog/june-6-sso-security-vulnerability - Record: https://forck.live/items/18094-june-6-2023-single-sign-on-security-vulnerability - Subject: Replit Agent Replit identified and patched a single-sign-on security vulnerability on June 6, 2023, after auditing its authentication systems. The vulnerability could have allowed an attacker to impersonate affected users via a fraudulent GitHub account under certain conditions, though Replit found no evidence of exploitation. As a precaution, Replit logged out all users who had ever used SSO with GitHub, Facebook, or Apple accounts. ### Hugging Face — The Hugging Face Hub for Galleries, Libraries, Archives and Museums - Company: Hugging Face (huggingface.co) - Announced: 2023-06-12 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/hf-hub-glam-guide - Record: https://forck.live/items/2045-the-hugging-face-hub-for-galleries-libraries-archives-and-museums - Subject: Platform GLAM professionals can use the Hugging Face Hub to find machine learning models, datasets, and demos, and how to contribute their own datasets. ### Hugging Face — Can foundation models label data like humans? - Company: Hugging Face (huggingface.co) - Announced: 2023-06-12 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/open-llm-leaderboard-rlhf - Record: https://forck.live/items/2044-can-foundation-models-label-data-like-humans - Subject: Platform This blog post investigates whether foundation models like GPT-4 can be used to evaluate other models' outputs by comparing their labels to human annotations. The authors curated a test set of prompts and completions from open-source models (Koala 13b, Vicuna 13b, OpenAssistant 12b, Dolly 12b) and collected both human preferences via Scale AI and GPT-4 evaluations. They expand the Hugging Face Open LLM Leaderboard to include automated benchmarks, human labels, and GPT-4 evaluations.