Qwen/Qwen3-TTS-12Hz-0.6B-CustomVoice
Text-to-Speech • 0.9B • Updated • 244k • 136
Sthenno
sthenno
AI & ML interests
To contact me: [email protected].
Recent Activity
liked a model 1 day ago
Qwen/Qwen3-TTS-12Hz-0.6B-CustomVoice liked a model 1 day ago
fishaudio/s2-pro liked a model 1 day ago
Qwen/Qwen3-ASR-0.6BOrganizations
liked 6
models 1 day ago
fishaudio/s2-pro
Text-to-Speech • 5B • Updated • 30.6k • 801
Qwen/Qwen3-ASR-0.6B
Automatic Speech Recognition • 0.9B • Updated • 438k • 260
Qwen/Qwen3-Reranker-0.6B
Text Ranking • 0.6B • Updated • 968k • 330
mudler/Qwen3.5-35B-A3B-APEX-GGUF
Text Generation • 35B • Updated • 24.3k • 50
amd/Qwen3.5-35B-A3B-MXFP4
Image-Text-to-Text • 21B • Updated • 72 • 1
upvoted a collection 1 day ago
liked 5
models 2 days ago
nvidia/Gemma-4-31B-IT-NVFP4
Text Generation • 21B • Updated • 44.2k • 165
google/gemma-4-31B
Image-Text-to-Text • 33B • Updated • 17.9k • 189
google/gemma-4-26B-A4B-it
Image-Text-to-Text • 27B • Updated • 133k • • 355
google/gemma-4-31B-it
Image-Text-to-Text • 33B • Updated • 287k • • 852
LiquidAI/LFM2.5-1.2B-Instruct-ONNX
Text Generation • Updated • 979 • 29
liked 3
models 3 days ago
upvoted a paper 3 days ago
repliedto SeaWolf-AI's post 3 days ago
Hello @SeaWolf-AI ,
This is a truly impressive piece of work. If you’re open to it, I would appreciate the opportunity to connect and explore potential collaboration.
You can find my email address in my profile—feel free to reach out. I look forward to hearing from you.
reactedto SeaWolf-AI's post with 🤗 3 days ago
Post
2113
🧬 Darwin-35B-A3B-Opus — The Child That Surpassed Both Parents
What if a merged model could beat both its parents? We proved it can.
Darwin-35B-A3B-Opus is a 35B MoE model (3B active) built with our Darwin V5 engine — the first evolution system that CT-scans parent models before merging them.
🤗 Model: FINAL-Bench/Darwin-35B-A3B-Opus
The result speaks for itself: GPQA Diamond 90.0%, versus Father (Qwen3.5-35B-A3B) at 84.2% and Mother (Claude 4.6 Opus Distilled) at 85.0%. That's +6.9% over Father and +5.9% over Mother. Not a tradeoff — a genuine leap. Meanwhile, MMMLU sits at 85.0% (Father: 85.2%), multimodal is fully intact, and all 201 languages are preserved.
How? Model MRI changed everything. Traditional merging is guesswork. Darwin V4 added evolution. Darwin V5 added X-ray vision. Model MRI scans each parent layer by layer and discovers: Mother's L34–L38 is the reasoning engine (peak cosine distance), 50–65% of Mother's experts are dead (killed by text-only distillation), and Father is a healthy generalist with every expert alive. The prescription: transplant Mother's reasoning brain at L38 (90% weight), replace her dead experts with Father's living ones, and let Father's router handle the output layer. Reasoning went up. Versatility stayed intact. No tradeoff — just evolution.
35B total, 3B active (MoE) · GPQA Diamond 90.0% · MMMLU 85.0% (201 languages) · Multimodal Image & Video · 262K native context · 147.8 tok/s on H100 · Runs on a single RTX 4090 (Q4) · Apache 2.0
Darwin V5's full algorithm and technical details will be released alongside an upcoming paper.
🚀 Live Demo: FINAL-Bench/Darwin-35B-A3B-Opus
🏆 FINAL Bench Leaderboard: FINAL-Bench/Leaderboard
📊 ALL Bench Leaderboard: FINAL-Bench/all-bench-leaderboard
Built by VIDRAFT · Supported by the Korean Government GPU Support Program
What if a merged model could beat both its parents? We proved it can.
Darwin-35B-A3B-Opus is a 35B MoE model (3B active) built with our Darwin V5 engine — the first evolution system that CT-scans parent models before merging them.
🤗 Model: FINAL-Bench/Darwin-35B-A3B-Opus
The result speaks for itself: GPQA Diamond 90.0%, versus Father (Qwen3.5-35B-A3B) at 84.2% and Mother (Claude 4.6 Opus Distilled) at 85.0%. That's +6.9% over Father and +5.9% over Mother. Not a tradeoff — a genuine leap. Meanwhile, MMMLU sits at 85.0% (Father: 85.2%), multimodal is fully intact, and all 201 languages are preserved.
How? Model MRI changed everything. Traditional merging is guesswork. Darwin V4 added evolution. Darwin V5 added X-ray vision. Model MRI scans each parent layer by layer and discovers: Mother's L34–L38 is the reasoning engine (peak cosine distance), 50–65% of Mother's experts are dead (killed by text-only distillation), and Father is a healthy generalist with every expert alive. The prescription: transplant Mother's reasoning brain at L38 (90% weight), replace her dead experts with Father's living ones, and let Father's router handle the output layer. Reasoning went up. Versatility stayed intact. No tradeoff — just evolution.
35B total, 3B active (MoE) · GPQA Diamond 90.0% · MMMLU 85.0% (201 languages) · Multimodal Image & Video · 262K native context · 147.8 tok/s on H100 · Runs on a single RTX 4090 (Q4) · Apache 2.0
Darwin V5's full algorithm and technical details will be released alongside an upcoming paper.
🚀 Live Demo: FINAL-Bench/Darwin-35B-A3B-Opus
🏆 FINAL Bench Leaderboard: FINAL-Bench/Leaderboard
📊 ALL Bench Leaderboard: FINAL-Bench/all-bench-leaderboard
Built by VIDRAFT · Supported by the Korean Government GPU Support Program
- 8 replies
liked 2
models 4 days ago
