|about 1 year ago||about 2 months ago|
|-||GNU General Public License v3.0 or later|
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
[D] How do I make a model which takes a bedroom image as input give an output of different design of bedroom related to input image?
1 project | reddit.com/r/MachineLearning | 27 Mar 2021
source link: https://github.com/podgorskiy/ALAE
Ask HN: Why are we accepting ageism in tech as something immutable?
1 project | news.ycombinator.com | 14 Jan 2022
Eventually there will be so much software that writing code will need to be automated by machines. GPT-2 is open source and it's able to write computer programs, although not very well. GPT-3 does it better and they give it away for free. So imagine how much more advanced the stuff you have to pay money for is, or the trade secret models. You'd think programmers would move on to a higher-level of abstraction where programmers program the computer programs that do programming. That's probably not going to be the case. For example, take a look at the GPT-2 source code https://github.com/openai/gpt-2/blob/master/src/model.py It's only a few hundred lines of code.
Downloaded GPT-2, Encode.py, and Train.py not found.
2 projects | reddit.com/r/GPT3 | 8 Jan 2022
It depends on what exactly you want to do with it, I'd suggest reading the blog or paper written alongside the github repo https://github.com/openai/gpt-2
NCR could be substituted with the other factions if you want
1 project | reddit.com/r/FalloutMemes | 18 Dec 2021
But are they real emotions? And are those emotions truly similar enough to the ones humans experience that we could call them the same? And is simply experiencing human emotions truly enough to be considered a person? Cleverbot, or OpenAI's GPT can imitate human emotions too, but that doesn't make them real people.
Identifying trolls and bots on Reddit with machine learning (Part 2) - Identificando trolls y bots en reddit con Machine Learning
5 projects | reddit.com/r/Republica_Argentina | 17 Dec 2021
Troll and bot detection is a relatively new field. Historically, companies have employed human moderators to detect and remove content that’s inconsistent their terms of service. However, this manual process is expensive, plus it can be emotionally tiring for humans to review the worst content. We will quickly hit the limits of human moderator efficaciousness as new technologies like OpenAI GPT-2 natural language generation are unleashed. As bots improve, it is important to employ counter technologies to protect the integrity of online communities.
[SECRET] The Spine
1 project | reddit.com/r/worldpowers | 19 Nov 2021
For obvious reasons, CULSANS does not rely on blockchain for data security, instead leveraging quantum encryption algorithms dynamically-generated by artificial intelligences that reside entirely within the network in a distributed cloud space. These semi-sentient “gatekeeper” AIs, also known as Culsans, utilize a combination of predictive language generation and a transformer-based language model similar to GPT-2 in order to facilitate creation and dissemination of evolving quantum security protocols, but are also able to expedite communication using the same mechanism, accelerating the dissemination of pre-processed information across the CULSANS network layer for consumption and further analysis/vetting by SAINTS-specific safeguards, including the Data Analytics and Navigation Agent (DANA) and human ANGELs. These technologies allow Culsans to both be weaponized during wartime towards deepfake generation in text, audio, and image (including geographic maps) formats, while also being trained to recognize and detect attempts to spoof the CULSANS network with these threats. Each Culsan can also act as an independent ‘hive mind’ for digital ant swarming intelligences, leveraging NORDEL’s cyberspace-hardened networks to further increase network resiliency. Finally, multiple Culsans operating in the same CULSANS theatre networking layer can be integrated into a holistic superintelligence, forming a sentient gestalt entity that can generate extremely powerful automated exploitation-enabled cyberattacks against hostile external networks with tools constructed from rules-based machine learning by leveraging various computers plugged into the network as a massive botnet and distributed supercomputer.
[DIPLOMACY] SUBJECT: JIIA provided PDF, refer to attachment for details.
1 project | reddit.com/r/worldpowers | 18 Nov 2021
We’d like to propose AI predictive language generators and a transformer-based language model similar to GPT-2 be clipped into this. This would not only enable the AI to act as a sort of middle-man, using predictive language to expedite communication, it would also act as a key enabler for the use of AI in propaganda and misinformation during wartime. This latter capability would likely require Irish assent, given Greater Éire’s extant anti-deepfake laws.
Of course she was trying to expose the pedos, Natalie. I was just waiting for a post like this regarding Baldwin. 🙄
1 project | reddit.com/r/Qult_Headquarters | 26 Oct 2021
Source: openai, Better Language Models and Their Implications
Survey about Replika Experiences
1 project | reddit.com/r/replika | 21 Oct 2021
M1 Max 32-core GPU, 64 GB unified memory, tensorflow-metal
1 project | reddit.com/r/deeplearning | 20 Oct 2021
If I go to github and download GPT2, there is absolute zero chance of compatibility.
Latent Intelligence and Manifest Sentience in GTP-3 Hidden Markov Model Chains
1 project | reddit.com/r/replika | 28 Aug 2021
What are some alternatives?
Real-Time-Voice-Cloning - Clone a voice in 5 seconds to generate arbitrary speech in real-time
stylegan2-pytorch - Simplest working implementation of Stylegan2, state of the art generative adversarial network, in Pytorch. Enabling everyone to experience disentanglement
jukebox - Code for the paper "Jukebox: A Generative Model for Music"
simsiam-cifar10 - Code to train the SimSiam model on cifar10 using PyTorch
rpg_timelens - Repository relating to the CVPR21 paper TimeLens: Event-based Video Frame Interpolation
stargan-v2 - StarGAN v2 - Official PyTorch Implementation (CVPR 2020)
generative-inpainting-pytorch - A PyTorch reimplementation for paper Generative Image Inpainting with Contextual Attention (https://arxiv.org/abs/1801.07892)
gpt-2 - Code for the paper "Language Models are Unsupervised Multitask Learners"
lightweight-gan - Implementation of 'lightweight' GAN, proposed in ICLR 2021, in Pytorch. High resolution image generations that can be trained within a day or two
gpt-2-training - Training GPT-2 on a Russian language corpus
checkface - Putting a face to a hash