In a Training Loop 🔄

50 17 82

DedeProGames PRO

DedeProGames

AI & ML interests

Thinking and Agentic Finetuning

Recent Activity

new activity about 9 hours ago

OrionLLM/GRM-2.6-Plus:dataset?

upvoted a collection 2 days ago

Minecraft-Base

updated a model 2 days ago

OrionLLM/GRM-2.6-Plus

View all activity

Organizations

New activity in OrionLLM/GRM-2.6-Plus about 9 hours ago

dataset?

#1 opened 3 days ago by

Roman1111111

upvoted a collection 2 days ago

Minecraft-Base

Collection

Datasets for fine-tuning Minecraft base vision language models. • 4 items • Updated Apr 19, 2025 • 3

updated a model 2 days ago

OrionLLM/GRM-2.6-Plus

Image-Text-to-Text • 28B • Updated 2 days ago • 179 • 10

upvoted a collection 2 days ago

GRM-2.6

Collection

1 item • Updated 3 days ago • 2

New activity in mradermacher/model_requests 2 days ago

OrionLLM/GRM-2.6-Plus

#2250 opened 2 days ago by

DedeProGames

liked a model 3 days ago

OrionLLM/GRM-2.6-Plus

Image-Text-to-Text • 28B • Updated 2 days ago • 179 • 10

published a model 3 days ago

OrionLLM/GRM-2.6-Plus

Image-Text-to-Text • 28B • Updated 2 days ago • 179 • 10

New activity in huggingface/InferenceSupport 3 days ago

OrionLLM/GRM-2.6-Plus

🚀 1

#9464 opened 3 days ago by

DedeProGames

updated a collection 3 days ago

GRM-2.6

Collection

1 item • Updated 3 days ago • 2

liked 2 models 3 days ago

Qwen/Qwen3.5-27B

Image-Text-to-Text • 28B • Updated 3 days ago • 3.35M • • 963

mradermacher/Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-Reasoning-i1-GGUF

8B • Updated Jan 2 • 114k • 49

reacted to danielhanchen's post with 🔥 3 days ago

Post

5104

Qwen3.6-27B is out now! Run it locally on 18GB RAM. 💜

Qwen3.6-27B surpasses Qwen3.5-397B-A17B on all major coding benchmarks.

GGUFs to run: unsloth/Qwen3.6-27B-GGUF
Guide + MLX: https://unsloth.ai/docs/models/qwen3.6

upvoted a collection 3 days ago

Qwen3.6

Collection

4 items • Updated 4 days ago • 243

reacted to imnotkitty's post with 🔥👍❤️ 3 days ago

Post

3871

tencent/Hy3-preview is out: an open-weights MoE reasoning model.

✅ 295B total / 21B active / 256K context
✅ Fused fast-and-slow thinking in a single model
✅ First model trained on Hunyuan's rebuilt pretraining + RL infra (Feb → Apr)

Benchmarks:
👉 SWE-Bench Verified, Terminal-Bench 2.0, BrowseComp, WideSearch — competitive results, particularly strong on agentic tool use
👉 Top score on Tsinghua's 2026 Spring math PhD qualifying exam
👉 Strong context-learning and instruction-following on Tencent's CL-bench / CL-bench-Life

More details can be found in my article: https://huggingface.co/blog/imnotkitty/hy3-preview