Fine-tuning GPT-3.5-turbo for natural language to SQL
35 by saigal | 7 comments on Hacker News.
Thursday, August 31, 2023
Wednesday, August 30, 2023
Tuesday, August 29, 2023
New top story on Hacker News: I raced a homing pigeon against the Internet [video]
I raced a homing pigeon against the Internet [video]
22 by tambourine_man | 9 comments on Hacker News.
22 by tambourine_man | 9 comments on Hacker News.
Monday, August 28, 2023
Sunday, August 27, 2023
Saturday, August 26, 2023
Friday, August 25, 2023
Thursday, August 24, 2023
Wednesday, August 23, 2023
Tuesday, August 22, 2023
Monday, August 21, 2023
Sunday, August 20, 2023
New top story on Hacker News: Show HN: Talk to AI Models in Terminal
Show HN: Talk to AI Models in Terminal
5 by today072 | 0 comments on Hacker News.
Hi everyone, nice to meet you and I am a newcomer of HN. I have made a binary tool Aih that could communicate with Bard, ChatGPT, Claude, and Llama(HuggingChat) from the terminal. https://ift.tt/BHuEnRP Since CAPTCHA challenges and bots detecting have become increasingly difficult, I've changed my strategy from hacking the APIs to simulating a real browser's action. The tool first takes the logged-in cookies of Google, ChatGPT, Claude, and HuggingChat accounts from the real Chrome browser, then it opens an invisible instance of Chromium for communication, then displays the answers in terminal. I think it's useful especially when I am researching some topics and need to compare answers of those AI models at the same time. Feel free to test and welcome provide feedback!
5 by today072 | 0 comments on Hacker News.
Hi everyone, nice to meet you and I am a newcomer of HN. I have made a binary tool Aih that could communicate with Bard, ChatGPT, Claude, and Llama(HuggingChat) from the terminal. https://ift.tt/BHuEnRP Since CAPTCHA challenges and bots detecting have become increasingly difficult, I've changed my strategy from hacking the APIs to simulating a real browser's action. The tool first takes the logged-in cookies of Google, ChatGPT, Claude, and HuggingChat accounts from the real Chrome browser, then it opens an invisible instance of Chromium for communication, then displays the answers in terminal. I think it's useful especially when I am researching some topics and need to compare answers of those AI models at the same time. Feel free to test and welcome provide feedback!
Saturday, August 19, 2023
Friday, August 18, 2023
Thursday, August 17, 2023
New top story on Hacker News: Show HN: Shadeform – Single Platform and API for Provisioning GPUs
Show HN: Shadeform – Single Platform and API for Provisioning GPUs
9 by edgoode | 1 comments on Hacker News.
Hi HN, we are Ed, Zach, and Ronald, creators of Shadeform ( https://ift.tt/aCZQoMf ), a GPU marketplace to see live availability and prices across the GPU market, as well as to deploy and reserve on-demand instances. We have aggregated 8+ GPU providers into a single platform and API, so you can easily provision instances like A100s and H100s where they are available. From our experience working at AWS and Azure, we believe that cloud could evolve from all-encompassing hyperscalers (AWS, Azure, GCP) to specialized clouds for high-performance use cases. After the launch of ChatGPT, we noticed GPU capacity thinning across major providers and emerging GPU and HPC clouds, so we decided it was the right time to build a single interface for IaaS across clouds. With the explosion of Llama 2 and open source models, we are seeing individuals, startups, and organizations struggling to access A100s and H100s for model fine-tuning, training, and inference. This encouraged us to help everyone access compute and increase flexibility with their cloud infra. Right now, we’ve built a platform that allows users to find GPU availability and launch instances from a unified platform. Our long term goal is to build a hardwareless GPU cloud where you can leverage managed ML services to train and infer in different clouds, reducing vendor lock-in. We shipped a few features to help teams access GPUs today: - a “single plane of glass” for GPU availability and prices; - a “single control plane” for provisioning GPUs in any cloud through our platform and API; - a reservation system that monitors real time availability and launches GPUs as soon as they become available. Next up, we’re building multi-cloud load balanced inference, streamlining self hosting open source models, and more. You can try our platform at https://ift.tt/hDxOlF3 . You can provision instances in your accounts by adding your cloud credentials and api keys, or you can leverage “ShadeCloud” and provision GPUs in our accounts. If you deploy in your account, it is free. If you deploy in our accounts, we charge a 5% platform fee. We’d love your feedback on how we’re approaching this problem. What do you think?
9 by edgoode | 1 comments on Hacker News.
Hi HN, we are Ed, Zach, and Ronald, creators of Shadeform ( https://ift.tt/aCZQoMf ), a GPU marketplace to see live availability and prices across the GPU market, as well as to deploy and reserve on-demand instances. We have aggregated 8+ GPU providers into a single platform and API, so you can easily provision instances like A100s and H100s where they are available. From our experience working at AWS and Azure, we believe that cloud could evolve from all-encompassing hyperscalers (AWS, Azure, GCP) to specialized clouds for high-performance use cases. After the launch of ChatGPT, we noticed GPU capacity thinning across major providers and emerging GPU and HPC clouds, so we decided it was the right time to build a single interface for IaaS across clouds. With the explosion of Llama 2 and open source models, we are seeing individuals, startups, and organizations struggling to access A100s and H100s for model fine-tuning, training, and inference. This encouraged us to help everyone access compute and increase flexibility with their cloud infra. Right now, we’ve built a platform that allows users to find GPU availability and launch instances from a unified platform. Our long term goal is to build a hardwareless GPU cloud where you can leverage managed ML services to train and infer in different clouds, reducing vendor lock-in. We shipped a few features to help teams access GPUs today: - a “single plane of glass” for GPU availability and prices; - a “single control plane” for provisioning GPUs in any cloud through our platform and API; - a reservation system that monitors real time availability and launches GPUs as soon as they become available. Next up, we’re building multi-cloud load balanced inference, streamlining self hosting open source models, and more. You can try our platform at https://ift.tt/hDxOlF3 . You can provision instances in your accounts by adding your cloud credentials and api keys, or you can leverage “ShadeCloud” and provision GPUs in our accounts. If you deploy in your account, it is free. If you deploy in our accounts, we charge a 5% platform fee. We’d love your feedback on how we’re approaching this problem. What do you think?
Wednesday, August 16, 2023
Tuesday, August 15, 2023
Monday, August 14, 2023
Sunday, August 13, 2023
Saturday, August 12, 2023
Friday, August 11, 2023
Thursday, August 10, 2023
Wednesday, August 9, 2023
Tuesday, August 8, 2023
Monday, August 7, 2023
New top story on Hacker News: Show HN: ChainForge, a visual tool for prompt engineering and LLM evaluation
Show HN: ChainForge, a visual tool for prompt engineering and LLM evaluation
11 by fatso784 | 0 comments on Hacker News.
Hi HN! We’re been working hard on this low-code tool for rapid prompt discovery, robustness testing and LLM evaluation. We’ve just released documentation to help new users learn how to use it and what it can already do. Let us know what you think! :)
11 by fatso784 | 0 comments on Hacker News.
Hi HN! We’re been working hard on this low-code tool for rapid prompt discovery, robustness testing and LLM evaluation. We’ve just released documentation to help new users learn how to use it and what it can already do. Let us know what you think! :)
Sunday, August 6, 2023
Saturday, August 5, 2023
Friday, August 4, 2023
Thursday, August 3, 2023
New top story on Hacker News: Tell HN: I think I found Toyota's battery
Tell HN: I think I found Toyota's battery
161 by scythe | 23 comments on Hacker News.
Recently there was a thread about a "breakthrough" in battery technology at Toyota. https://ift.tt/aJyLrCn Toyota has been putting out PR puff pieces about their "solid-state" (solid-electrolyte) batteries for years, but this story was unique in that it had a quote from Keiji Kaita, who holds some high-level role at Toyota. Anyway, I didn't think much of it, because there was no paper referenced in the Guardian article, which seemed to be the original source. But while reading about something else, I came across the paper "A near dimensionally invariable high-capacity positive electrode material", published in Nature Materials last December: https://ift.tt/AuW7UQF This paper, reporting a cathode that has very little (much less than normal) change in size or shape when charged and discharged, claims reversible storage with a solid electrolyte. It stands to reason that dimensional stability of the cathode is necessary for interfacing with a solid electrolyte, since if it swells and shrinks, it will probably detach from the electrolyte, and possibly damage it further. Looking at the affiliations of some of the authors we see a number of contributors from the "Lithium Ion Battery Technology and Evaluation Center (LIBTEC)". A web search about LIBTEC leads to several articles from 2018: https://ift.tt/zPShvWM... which state that Toyota, along with Nissan, Honda and Panasonic (Tesla's major collaborator), have established this consortium to work on solid-electrolyte batteries as of five years ago. So what does this thing look like? It's a vanadium–titanium cathode, Li8Ti2V4O14. Titanium is common; vanadium technically has a higher crustal abundance than nickel, but it tends to be spread across low-quality deposits, so production is low right now. A review considering the resource outlook for V-based batteries [1] was guardedly optimistic. 750 Wh/kg is great . Vanadium cathodes historically had a problem with high dimensional instability , but it appears that cocrystallization with titanium may have fixed that, and the weird properties of vanadium became an advantage in compensating for Li+ influx/efflux. The use of a sulfide electrolyte pours doubt on claims of safety, though. It's reasonably likely that if water were to come into contact with the electrolyte, it could release highly toxic hydrogen sulfide gas. Also, since the battery was developed in collaboration with other major automakers (and funded by the Japanese government), it's somewhat questionable to think it would give Toyota a major advantage in the EV race. But for the Japanese economy, which has been rather slow lately, it could be a boost. 1: https://ift.tt/Ly0TKJe....
161 by scythe | 23 comments on Hacker News.
Recently there was a thread about a "breakthrough" in battery technology at Toyota. https://ift.tt/aJyLrCn Toyota has been putting out PR puff pieces about their "solid-state" (solid-electrolyte) batteries for years, but this story was unique in that it had a quote from Keiji Kaita, who holds some high-level role at Toyota. Anyway, I didn't think much of it, because there was no paper referenced in the Guardian article, which seemed to be the original source. But while reading about something else, I came across the paper "A near dimensionally invariable high-capacity positive electrode material", published in Nature Materials last December: https://ift.tt/AuW7UQF This paper, reporting a cathode that has very little (much less than normal) change in size or shape when charged and discharged, claims reversible storage with a solid electrolyte. It stands to reason that dimensional stability of the cathode is necessary for interfacing with a solid electrolyte, since if it swells and shrinks, it will probably detach from the electrolyte, and possibly damage it further. Looking at the affiliations of some of the authors we see a number of contributors from the "Lithium Ion Battery Technology and Evaluation Center (LIBTEC)". A web search about LIBTEC leads to several articles from 2018: https://ift.tt/zPShvWM... which state that Toyota, along with Nissan, Honda and Panasonic (Tesla's major collaborator), have established this consortium to work on solid-electrolyte batteries as of five years ago. So what does this thing look like? It's a vanadium–titanium cathode, Li8Ti2V4O14. Titanium is common; vanadium technically has a higher crustal abundance than nickel, but it tends to be spread across low-quality deposits, so production is low right now. A review considering the resource outlook for V-based batteries [1] was guardedly optimistic. 750 Wh/kg is great . Vanadium cathodes historically had a problem with high dimensional instability , but it appears that cocrystallization with titanium may have fixed that, and the weird properties of vanadium became an advantage in compensating for Li+ influx/efflux. The use of a sulfide electrolyte pours doubt on claims of safety, though. It's reasonably likely that if water were to come into contact with the electrolyte, it could release highly toxic hydrogen sulfide gas. Also, since the battery was developed in collaboration with other major automakers (and funded by the Japanese government), it's somewhat questionable to think it would give Toyota a major advantage in the EV race. But for the Japanese economy, which has been rather slow lately, it could be a boost. 1: https://ift.tt/Ly0TKJe....