# forck.live — 31 January – 6 February 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-W05, 31 January – 6 February 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-W05 - The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md - Catalogue: https://forck.live/llms.txt ## Announcements ### Replit — Making new Python repls 100x faster to start up - Company: Replit (replit.com) - Announced: 2022-02-04T19:45:00+00:00 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://replit.com/blog/python-new-template - Record: https://forck.live/items/18084-making-new-python-repls-100x-faster-to-start-up - Subject: Replit Agent Replit released a new Python template that uses a content-addressable cache of individual package files, allowing symbolic links instead of full file copies. This reduces disk usage so large packages like TensorFlow fit on repls, and makes hosted Python repls start faster because the needed files are already present. ### Replit — What We Look for When We Interview - Company: Replit (replit.com) - Announced: 2022-02-03 - Category: not stated - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://replit.com/blog/get-hired - Record: https://forck.live/items/18166-what-we-look-for-when-we-interview - Subject: Replit Agent We’re on a mission to bring the next billion software creators online. In 2022, to get there, we're focusing on: Making Replit run fast on popular devices across the globe. ### OpenAI — Solving (some) formal math olympiad problems - Company: OpenAI (openai.com) - Announced: 2022-02-02T08:00:00+00:00 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://openai.com/index/formal-math - Record: https://forck.live/items/917-solving-some-formal-math-olympiad-problems - Subject: GPT / ChatGPT / API OpenAI built a neural theorem prover for Lean that solves challenging high-school olympiad problems, including AMC12, AIME, and IMO problems. ### Hugging Face — Getting Started with Sentiment Analysis using Python - Company: Hugging Face (huggingface.co) - Announced: 2022-02-02 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/sentiment-analysis-python - Record: https://forck.live/items/2226-getting-started-with-sentiment-analysis-using-python - Subject: Platform A tutorial guide on sentiment analysis using Python, covering pre-trained models from Hugging Face Hub and building custom models. ### Hugging Face — Making automatic speech recognition work on large files with Wav2Vec2 in 🤗 Transformers - Company: Hugging Face (huggingface.co) - Announced: 2022-02-01 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/asr-chunking - Record: https://forck.live/items/2227-making-automatic-speech-recognition-work-on-large-files-with-wav2vec2-in - Subject: Platform To use chunking with stride to perform automatic speech recognition on arbitrarily long files using Wav2Vec2 in the Transformers library, leveraging the CTC architecture to maintain quality. ### Replit — Understanding Repl Resource Utilization - Company: Replit (replit.com) - Announced: 2022-01-31T20:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://replit.com/blog/repl-resources - Record: https://forck.live/items/18061-understanding-repl-resource-utilization - Subject: Replit Agent Replit's guide explains how CPU, RAM, and storage are allocated to Repls: free plans get 0.5 vCPUs, 1 GiB memory, and 1 GB SSD storage; Hacker plans get 2 vCPUs and 2 GB memory (4 GB when boosted); Boosted repls get 4 vCPUs. The post describes a new resource monitoring component in the file tree to help users track utilization.