meta-models/Muse-Glimmer-30B
Image-Text-to-Text • 30B • Updated • • 1.31k
enzo PRO
enzostvs
AI & ML interests
crafting stuffs at hugging face
Recent Activity
liked a model 1 day ago
meta-models/Muse-Glimmer-30B upvoted a collection 6 days ago
Hub 📊 liked a Space 8 days ago
multimodalart/minimax-h3Organizations
liked a
model 1 day ago
upvoted a collection 6 days ago
liked a
Space 8 days ago
liked a
model 8 days ago
upvoted a changelog 9 days ago
Hugging Face Changelog
Filter Jobs by Label
• 87
liked a
Space 14 days ago
liked a
Space 17 days ago
liked a
model 18 days ago
liked a
model 21 days ago
upvoted 2
changelogs 22 days ago
Hugging Face Changelog
MCP Server Enhancements
• 199
Hugging Face Changelog
Egress metrics for users and organizations
• 113
test new discussion
1
#569 opened 27 days ago
by enzostvs
upvoted a changelog 28 days ago
Hugging Face Changelog
Build Spaces with AI Agents
• 147
liked a
Space 29 days ago
upvoted a changelog 29 days ago
Hugging Face Changelog
Fine-Grained Token Presets
• 73
liked a
Space about 1 month ago
reacted to Quazim0t0's post with 🚀 about 1 month ago
Post
5157
🌼 DaisyChain-Web: train a language model with friends or by yourself with multiple devices, in the browser, no install
Open a webpage, share a room link, and every device that joins becomes part of the training cluster. Phones, laptops, old PCs: they connect peer-to-peer over WebRTC and train one shared transformer together, entirely in the browser.
What's actually happening under the hood:
🧠 A mini transformer LM trains on FineWeb-Edu, streamed live from the HuggingFace Hub. Each device pulls its own slice (data parallelism), tokenized with our 16.5k-token Spikewhale tokenizer
⚡ Every single multiply runs through verified INT8 neural units, no float fallback. On WebGPU browsers it uses the GPU's DP4A integer dot-product hardware, admitted only after proving bit-identical results against the verified units, with a 3×INT8 fast-accurate scheme (CUTLASS's 3xTF32 trick, ported to 8-bit)
🔒 Devices average gradients every step under a sync guard: a per-step roster protocol plus weight-hash verification keeps every device's model bit-identical. If anything drifts, training stops instead of silently forking
📊 Live logs show exactly what every device contributes, step by step
💾 When you're done: test generations right on the page, download a checkpoint, or grab the inference kit, a single self-contained HTML file with the weights baked in that runs generations offline, anywhere
Works solo too. Every extra device just grows the effective batch.
👉 Try it: Quazim0t0/DaisyChain-Web
🛠 Training framework: DaisyChainAI/DaisyChain-Train
Updates:
- Block-scaled INT8 quantization
- Batched attention GEMM
- Fused dequant+ReLU epilogue
- Weight-tied unembedding
- WebSocket relay fallback
- Server keepalive ping/pong every 30s
- disconnected-state redial
- Visibility/network-change reconnect
Open a webpage, share a room link, and every device that joins becomes part of the training cluster. Phones, laptops, old PCs: they connect peer-to-peer over WebRTC and train one shared transformer together, entirely in the browser.
What's actually happening under the hood:
🧠 A mini transformer LM trains on FineWeb-Edu, streamed live from the HuggingFace Hub. Each device pulls its own slice (data parallelism), tokenized with our 16.5k-token Spikewhale tokenizer
⚡ Every single multiply runs through verified INT8 neural units, no float fallback. On WebGPU browsers it uses the GPU's DP4A integer dot-product hardware, admitted only after proving bit-identical results against the verified units, with a 3×INT8 fast-accurate scheme (CUTLASS's 3xTF32 trick, ported to 8-bit)
🔒 Devices average gradients every step under a sync guard: a per-step roster protocol plus weight-hash verification keeps every device's model bit-identical. If anything drifts, training stops instead of silently forking
📊 Live logs show exactly what every device contributes, step by step
💾 When you're done: test generations right on the page, download a checkpoint, or grab the inference kit, a single self-contained HTML file with the weights baked in that runs generations offline, anywhere
Works solo too. Every extra device just grows the effective batch.
👉 Try it: Quazim0t0/DaisyChain-Web
🛠 Training framework: DaisyChainAI/DaisyChain-Train
Updates:
- Block-scaled INT8 quantization
- Batched attention GEMM
- Fused dequant+ReLU epilogue
- Weight-tied unembedding
- WebSocket relay fallback
- Server keepalive ping/pong every 30s
- disconnected-state redial
- Visibility/network-change reconnect
- 11 replies
liked a
Space about 1 month ago
updated a
dataset about 1 month ago
published a
dataset about 1 month ago