# forck.live — 10–16 January 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-W02, 10–16 January 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-W02 - The current issue, and the catalogue of the other weeks: https://forck.live/briefing.md - Catalogue: https://forck.live/llms.txt ## Announcements ### Hugging Face — Case Study: Millisecond Latency using Hugging Face Infinity and modern CPUs - Company: Hugging Face (huggingface.co) - Announced: 2022-01-13 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/infinity-cpu-performance - Record: https://forck.live/items/2230-case-study-millisecond-latency-using-hugging-face-infinity-and-modern-cpus - Subject: Platform Performance benchmarks of Hugging Face Infinity on Intel Xeon CPUs, showing up to 34% better latency/throughput compared to previous generation and up to 800% better than vanilla transformers. It also notes that Infinity is no longer offered as a commercial inference solution as of December 2022. ### Hugging Face — Boosting Wav2Vec2 with n-grams in 🤗 Transformers - Company: Hugging Face (huggingface.co) - Announced: 2022-01-12 - Category: capability-change - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://huggingface.co/blog/wav2vec2-with-ngram - Record: https://forck.live/items/2231-boosting-wav2vec2-with-n-grams-in-transformers - Subject: Platform Hugging Face Transformers now integrates with Kensho Technologies' pyctcdecode library to allow easy n-gram language model decoding with fine-tuned Wav2Vec2 checkpoints, improving transcription accuracy. ### Runway — Soundify: Matching sound effects to video - Company: Runway (runway.com) - Announced: 2022-01-11T22:24:14.964000+00:00 - Category: research-paper - Coverage: not counted - Announcement: yes - Group: announcements - Source: https://runway.com/research/soundify-matching-sound-effects-to-video - Record: https://forck.live/items/7944-soundify-matching-sound-effects-to-video - Subject: Gen / Aleph Runway and Carnegie Mellon University introduce Soundify, a system that matches sound effects to video by extending CLIP into a zero-shot detector, addressing bottlenecks in finding, aligning, and tuning sounds for video editing. ### Replit — Migrating our Web App from Heroku to GCP - Company: Replit (replit.com) - Announced: 2022-01-11T20:00:00+00:00 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://replit.com/blog/heroku2gcp - Record: https://forck.live/items/18169-migrating-our-web-app-from-heroku-to-gcp - Subject: Replit Agent Replit has fully migrated its web application from Heroku to Google Cloud Platform. The migration involved moving Postgres and Redis databases to GCP, then switching the front-end using Cloudflare Load Balancer to gradually redirect traffic. The process required multiple maintenance windows and extensive rehearsals to ensure a seamless transition for users. ### Hugging Face — Deploy GPT-J 6B for inference using Hugging Face Transformers and Amazon SageMaker - Company: Hugging Face (huggingface.co) - Announced: 2022-01-11 - Category: not stated - Coverage: not counted - Announcement: no - Group: routine - Source: https://huggingface.co/blog/gptj-sagemaker - Record: https://forck.live/items/2232-deploy-gpt-j-6b-for-inference-using-hugging-face-transformers-and-amazon - Subject: Platform Deploying the GPT-J 6B model for inference using Hugging Face Transformers and Amazon SageMaker, focusing on reducing model loading time by using torch.save and torch.load to achieve production-compatible performance.