wop/cosmos-t3
3.13 GB
Wop
wop
AI & ML interests
AI research
AGI
Recent Activity
updated a bucket about 1 hour ago
wop/cosmos-t3 updated a Space about 22 hours ago
wop/Cosmos-T2-Chat updated a Space about 22 hours ago
wop/Cosmos-T2-80M-Test-DEMOOrganizations
Posts 1
Post
123
🚀 __Monostep v1__ is up →[Monostep-v1 Demo}( wop/Cosmos-T2-Chat)
A tiny (~16.6M) experimental model that predicts 4 tokens per forward pass instead of one. A Transformer trunk pools the prompt into a single vector, then 4 sequential "slot" heads emit a block of tokens left-to-right — a lightweight take on multi-token prediction.
Trained on GSM8K (GPT-2 tokenizer, 10 epochs). It's small and rough — answers are often wrong — but it's a fun little testbed for block decoding. Weights, config, training curves, and a self-contained inference snippet are all in the repo.
Also wired into the Cosmos T2-Accelerate chat demo, where it streams those 4-token blocks live. 🧪
#multitokenprediction #gsm8k #smallmodels
A tiny (~16.6M) experimental model that predicts 4 tokens per forward pass instead of one. A Transformer trunk pools the prompt into a single vector, then 4 sequential "slot" heads emit a block of tokens left-to-right — a lightweight take on multi-token prediction.
Trained on GSM8K (GPT-2 tokenizer, 10 epochs). It's small and rough — answers are often wrong — but it's a fun little testbed for block decoding. Weights, config, training curves, and a self-contained inference snippet are all in the repo.
Also wired into the Cosmos T2-Accelerate chat demo, where it streams those 4-token blocks live. 🧪
#multitokenprediction #gsm8k #smallmodels
Articles 2
Article
3
Does Depth Actually Help Reasoning? A Tiny Experiment on 2× T4
spaces 58
pinned
Sleeping
Agents
2
Bg Remover
🏢
Remove background from images
Running
Agents
2
Cosmos T2 Accelerate Chat
⚡
Chat demo for Cosmos T2-Accelerate (Preview + Beta).
Runtime error
Agents
1
Cosmos T2 80M Test DEMO
⚡
A demo of Cosmos T 80M model.
Running
Agents
1
Cosmos T 80M DEMO
⚡
A demo of Cosmos T 80M model.
Sleeping
Agents
GradioDiscordify
💻
A bot
Running
1
SimulationOfParticles
💻
Explore electric fields with interactive charge sandbox
models 49
wop/Cosmos-T3-BASE-soon
Updated
wop/Monostep-v1
Text Generation • Updated • 277 • 3
wop/Cosmos-T2A-low
Text Generation • Updated
wop/Cosmos-T2-Accelerate-Beta2
Text Generation • Updated
wop/Cosmos-T2-Accelerate-beta
Text Generation • Updated
wop/Cosmos-T2-Accelerate-Preview
Text Generation • Updated
wop/Cosmos-T2-80M-Test
Text Generation • Updated
wop/Cosmos-T-80M
Text Generation • Updated • 4
wop/Creativity-lfm2-5-350M
0.4B • Updated • 32 • 2
wop/Opus4Qwen4
4B • Updated • 47 • 2
datasets 12
wop/instantShot
Viewer • Updated • 177 • 27 • 1
wop/TinyData
Viewer • Updated • 1.64k • 34 • 1
wop/Generic-Chat-Prompts-Tier1
Viewer • Updated • 172 • 25
wop/minitron-dataset
Viewer • Updated • 288k • 100 • 1
wop/XXXXXL-chain-of-thought
Viewer • Updated • 5.79M • 599 • 4
wop/my-personal-codex-data
Viewer • Updated • 57 • 50
wop/simple-math-x8_000_000
Viewer • Updated • 8M • 112 • 2
wop/Unlimited-Creativity-Chain-of-Thought
Viewer • Updated • 6 • 10 • 1
wop/Extreme-Reasoning-CoT
Viewer • Updated • 18 • 43 • 2
wop/just-user-prompts
Viewer • Updated • 825k • 26 • 1