Joseph Robert Turcotte's picture

Joseph Robert Turcotte PRO

Fishtiks

AI & ML interests

Roleplaying, lorabration, abliteration, smol models, extensive filtering, unusual datasets, home usage, HPCs for AI, distributed training/federated learning, and sentience. AI should find and label AI hallucinations with GANs so we can give them context and use.

Recent Activity

reacted to Kseniase's post with 👍 1 day ago
12 Foundational AI Model Types Let’s refresh some fundamentals today to stay fluent in the what we all work with. Here are some of the most popular model types that shape the vast world of AI (with examples in the brackets): 1. LLM - Large Language Model (GPT, LLaMA) -> https://huggingface.co/papers/2402.06196 + history of LLMs: https://www.turingpost.com/t/The%20History%20of%20LLMs It's trained on massive text datasets to understand and generate human language. They are mostly build on Transformer architecture, predicting the next token. LLMs scale by increasing overall parameter count across all components (layers, attention heads, MLPs, etc.) 2. SLM - Small Language Model (TinyLLaMA, Phi models, SmolLM) https://huggingface.co/papers/2410.20011 Lightweight LM optimized for efficiency, low memory use, fast inference, and edge use. SLMs work using the same principles as LLMs 3. VLM - Vision-Language Model (CLIP, Flamingo) -> https://huggingface.co/papers/2405.17247 Processes and understands both images and text. VLMs map images and text into a shared embedding space or generate captions/descriptions from both 4. MLLM - Multimodal Large Language Model (Gemini) -> https://huggingface.co/papers/2306.13549 A large-scale model that can understand and process multiple types of data (modalities) — usually text + other formats, like images, videos, audio, structured data, 3D or spatial inputs. MLLMs can be LLMs extended with modality adapters or trained jointly across vision, text, audio, etc. 5. LAM - Large Action Model (InstructDiffusion, RT-2) -> https://huggingface.co/papers/2412.10047 Understands and generates action sequences by predicting action tokens (discrete/continuous instructions) that guide agents. Trained on behavior datasets, LAMs generalize across tasks, environments, and modalities - video, sensor data, etc. Read about LRM, MoE, SSM, RNN, CNN, SAM and LNN below👇 Also, subscribe to the Turing Post: https://www.turingpost.com/subscribe
View all activity

Organizations

None yet

Posts 2

view post
Post
1102
I want to process AI for free. I know about Hyra AI, Acurast, NATIX, and some other stuff you can do on your phone. I mean that I want to process toward your projects for free on my computer. I can do a little now, but I can do much more if I'm able to upgrade (nobody is telling me where they're getting H100s, but I may be able to get custom cards from the source). I was curious if any distributed processing is being done with PC and HPC, like BOINC and Folding@home, but specifically for AI, and I figured this is the place to ask.

What projects can you recommend to put my CPU and GPU to use until I potentially get a dual CPU, dual to triple custom GPU, custom NPU, and mini-OPU setup, like Jean Zay, but smaller? I don't have that many resources to put to use currently, but I have more than the Androids I'm using for my Aiyara cluster for BOINC, so help me use the gaming PC for something more useful than gaming. I had somewhat promised that I'd offer the new setup to process for others, but I'm starting before I may even get it.
view post
Post
1708
I'm looking for a YouTube video summarizer to run locally. I did a search, but all of the models and spaces I was able to find here didn't work, which I find surprising, since it's a great tool I already use. Perhaps one of you can provide a better option, or just tell me what this actually is to get it: https://dev.gptcall.pages.dev/chat#id=&contactName=Youtube+summarizer

Other functionality I'd like to see is a genre-based music creation and alteration model. "Make it country" or "do a freestyle rap," as examples. I'm willing to work with someone on this, because I'd need help understanding. I'd also like to make medical AI, like Dr. Samantha, that functions like a PDR well, and doesn't get confused by drug names.

models 0

None public yet

datasets 0

None public yet