Posts

Showing posts from August, 2023

WizardCoder 34B based on Meta's Code Llama 2 (Beats GPT4 on HumanEval with 73.2% pass@1) #1208

Image
WizardCoder 34B based on Meta's Code Llama 2 (Beats GPT4 on HumanEval with 73.2% pass@1) #1208 Source  

Apptronik is building humanoid robot - starting with Apollo (and in collab with NASA) #1207

Image
Apptronik is building humanoid robot - starting with Apollo (and in collab with NASA) #1207 Source  

Agibot RAISE-A1 - new robot prototype with new robotics company #1206

Image
Agibot RAISE-A1 - new robot prototype with new robotics company #1206 Source  

phind - Beating GPT-4 on HumanEval with a Fine-Tuned CodeLlama-34B (better than GPT-4 on code and open source) #1205

Image
phind - Beating GPT-4 on HumanEval with a Fine-Tuned CodeLlama-34B (better than GPT-4 on code and open source) #1205 Source  

Meta: Introducing Code Llama, a state-of-the-art large language model for coding (Based on Llama 2) #1204

Image
Meta: Introducing Code Llama, a state-of-the-art large language model for coding (Based on Llama 2) #1204 Source  

Giraffe - Long Context Open-Source LLMs (32k context based on Llama 2) #1203

Image
Giraffe - Long Context Open-Source LLMs (32k context based on Llama 2) #1203 Source

Introducing Eleven Multilingual v2: our new AI speech model supporting 28 languages #1202

Image
Introducing Eleven Multilingual v2: our new AI speech model supporting 28 languages #1202 Source  

OpenAI: GPT-3.5 Turbo fine-tuning and API updates #1201

Image
OpenAI: GPT-3.5 Turbo fine-tuning and API updates #1201 Source  

Meta AI: Introducing SeamlessM4T, the first all-in-one, multilingual multimodal translation model #1200

Image
Meta AI: Introducing SeamlessM4T, the first all-in-one, multilingual multimodal translation model #1200 Source  

Dolma - The largest open source text dataset just dropped (3T Tokens) #1199

Image
Dolma - The largest open source text dataset just dropped (3T Tokens) #1199 Source  

Meta's RoboAgent can acquire a wide diversity of non-trivial skills + generalize them to hundreds of unseen scenarios #1198

Image
Meta's RoboAgent can acquire a wide diversity of non-trivial skills + generalize them to hundreds of unseen scenarios #1198 Source

Meta AI's Shepherd is a language model specifically tuned to critique model responses & suggest refinements #1197

Image
Meta AI's Shepherd is a language model specifically tuned to critique model responses & suggest refinements #1197 Source  

Introducing Unitree H1: Its First General-purpose Humanoid Robot| Embodied AI, Price below $90k #1196

Image
Introducing Unitree H1: Its First General-purpose Humanoid Robot| Embodied AI, Price below $90k #1196 Source  

Using GPT-4 for content moderation #1195

Image
Using GPT-4 for content moderation #1195 Source  

Runway AI now allows up to 16 seconds of videos #1194

Image
Runway AI now allows up to 16 seconds of videos #1194 Source  

AI Is Building Highly Effective Antibodies That Humans Can’t Even Imagine #1193

Image
AI Is Building Highly Effective Antibodies That Humans Can’t Even Imagine #1193 Source  

Nvidia unveils more powerful AI chip coming next year (GH200 with 3x memory of H100) #1192

Image
Nvidia unveils more powerful AI chip coming next year (GH200 with 3x memory of H100) #1192 Source  

A long wave of Luddism is coming against AI progress #1191

Image
A long wave of Luddism is coming against AI progress #1191 Source  

OpenAI's ChatGPT Custom instructions are now available to ChatGPT users on the free plan #1190

Image
OpenAI's ChatGPT Custom instructions are now available to ChatGPT users on the free plan #1190 Source  

AnthropicAI: Introducing our latest version of Claude Instant, version 1.2, available now through our API #1189

Image
AnthropicAI: Introducing our latest version of Claude Instant, version 1.2, available now through our API #1189 Source  

Google: Introducing Project IDX, An Experiment to Improve Full-stack, Multiplatform App Development (powered by AI) #1188

Image
Google: Introducing Project IDX, An Experiment to Improve Full-stack, Multiplatform App Development (powered by AI) #1188 Source  

Hollywood vs AI: Disney looks to cut more costs by using artificial intelligence #1187

Image
Hollywood vs AI: Disney looks to cut more costs by using artificial intelligence #1187 Source  

AvatarVerse: High-quality & Stable 3D Avatar Creation from Text and Pose #1186

Image
AvatarVerse: High-quality & Stable 3D Avatar Creation from Text and Pose #1186 Source  

AGI is only 2-3 years away according to AnthropicAI's CEO #1185

Image
Source  

Stability AI has launched StableCode, the revolutionary generative AI LLM for coding #1184

Image
Stability AI has launched StableCode, the revolutionary generative AI LLM for coding #1184 Source  

LK-99 is likely simply a ferromagnetic material, not a superconductor (it's over) #1183

Image
LK-99 is likely simply a ferromagnetic material, not a superconductor (it's over) #1183 Source  

Life tips from reddit I should follow #1182

Image
Life tips from reddit I should follow #1182 So basically this is a collection of screenshots from reddit. Tips and stuff I should remember and follow.  (8 August 2023) (9 August 2023) (11 August 2023) (12 August 2023) (16 August 2023) (25 August 2023)

Being childfree in future, that's what I want and probably no marriage either #1181

Image
Being childfree in future, that's what I want and probably no marriage either #1181

Delivery robots in US are being attacked by people (this is so futuristic lol) #1180

Image
Delivery robots in US are being attacked by people (this is so futuristic lol) #1180 Source  

First fully levitating LK-99 Superconductor from China (unverified) #1179

Image
First fully levitating LK-99 Superconductor from China (unverified) #1179 Source   X

Alibaba Open Sources Qwen, a 7B Parameter AI Model #1178

Image
Alibaba Open Sources Qwen, a 7B Parameter AI Model #1178 Source  

Immersed Announces 'Visor' Compact Productivity Headset With 4K OLED Microdisplays #1177

Image
Immersed Announces 'Visor' Compact Productivity Headset With 4K OLED Microdisplays #1177 Source  

AudioLDM 2: text to audio/music/speech generation #1176

Image
AudioLDM 2: text to audio/music/speech generation #1176 Source  

Nvidia Perfusion - Image Generator Fits on a Floppy Disk and Takes 4 Minutes to Train #1175

Image
Nvidia Perfusion - Image Generator Fits on a Floppy Disk and Takes 4 Minutes to Train #1175 Source  

Meta AI open-sources AudioCraft which lets users create high quality music and audio #1174

Image
Meta AI open-sources AudioCraft which lets users create high quality music and audio #1174 Source

First claimed successful replication of LK-99 Superconductor #1173

Image
First claimed successful replication of LK-99 Superconductor #1173 Source  

National Lab (LBNL) results support LK-99 as a room-temperature ambient-pressure superconductor #1172

Image
National Lab (LBNL) results support LK-99 as a room-temperature ambient-pressure superconductor #1172 31 July  Source