-
meta-llama/Llama-3.1-70B
Text Generation • 71B • Updated • 86.3k • 410 -
Qwen/Qwen3.5-35B-A3B
Image-Text-to-Text • 36B • Updated • 2.93M • • 1.26k -
deepseek-ai/DeepSeek-R1
Text Generation • 685B • Updated • 1.98M • • 13.1k -
mistralai/Mistral-Large-3-675B-Instruct-2512
Updated • 667 • 219
Croc-Prog-HF
Croc-Prog-HF
AI & ML interests
High-temperature Text-Generation("Creativity" of the model) and Etical AI Alignment.
Democratizing local inference: reducing hardware requirements. Against the oligopoly of huge and polluting LLMs.
Recent Activity
upvoted a paper about 5 hours ago
OpenResearcher: A Fully Open Pipeline for Long-Horizon Deep Research Trajectory Synthesis reacted to omarkamali's post with 🤯 about 5 hours ago
I just might have cracked tokenizer-free LLMs. No vocab, no softmax.
I'm training a 22M params LLM rn to test this "thing" and it's able to formulate coherent sentences 🤯
Bear in mind, this is a completely new, tokenizer-free LLM architecture with built-in language universality.
Check the explainer video to understand what's happening. Feedback welcome on this approach!
updated a dataset about 23 hours ago
Croc-Prog-HF/Creative-knowledge-for-WritingOrganizations
None yet
-
meta-llama/Llama-3.1-70B
Text Generation • 71B • Updated • 86.3k • 410 -
Qwen/Qwen3.5-35B-A3B
Image-Text-to-Text • 36B • Updated • 2.93M • • 1.26k -
deepseek-ai/DeepSeek-R1
Text Generation • 685B • Updated • 1.98M • • 13.1k -
mistralai/Mistral-Large-3-675B-Instruct-2512
Updated • 667 • 219