wop/Cosmos-SFT
1.57 GB
Wop
wop
AI & ML interests
AI research
AGI
Recent Activity
liked a Space about 4 hours ago
wop/visualize new activity about 4 hours ago
SupraLabs/Blog:Discord server liked a Space about 5 hours ago
AxiomicLabs/Open_SLM_LeaderboardOrganizations
Posts 1
Post
182
🚀 __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 61
pinned
Running
1
Distill yourself
👀
Annotate text with emotion intensity and export as dataset
pinned
Running
4
Fontend
🚀
Chat with an AI assistant in real time
pinned
Sleeping
Agents
2
Bg Remover
🏢
Remove background from images
Running
1
Visualize
🚀
Explore and visualize emotions flowing through text entries
Running
Agents
1
Server Backend
⚡
Chat with the Cosmos‑T3 language model for instant responses
Paused
Agents
2
Cosmos T2 Accelerate Chat
⚡
Chat demo for Cosmos T2-Accelerate (Preview + Beta).
models 49
wop/Cosmos-T3-BASE-soon
Updated
wop/Monostep-v1
Text Generation • Updated • 606 • 4
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 • 25 • 2
wop/Opus4Qwen4
4B • Updated • 36 • 2
datasets 18
wop/Human-Essence-Dataset
Updated • 1
wop/debug-self
Viewer • Updated • 1
wop/smallestdata-sft
Viewer • Updated • 19 • 52 • 1
wop/Super-Intelligence-Immersive
Viewer • Updated • 2 • 30
wop/Cosmos-T3-06-09-2026
Viewer • Updated • 295k • 59 • 1
wop/miniconveserations
Viewer • Updated • 16 • 32
wop/instantShot
Viewer • Updated • 177 • 74 • 1
wop/TinyData
Viewer • Updated • 1.64k • 95 • 2
wop/Generic-Chat-Prompts-Tier1
Viewer • Updated • 172 • 21
wop/minitron-dataset
Viewer • Updated • 288k • 113 • 1