tencent/Hy3
Text Generation • 299B • Updated • 20.7k • • 887
Piotr Sobczak
Haloons
·
AI & ML interests
None yet
Recent Activity
liked a model 21 days ago
tencent/Hy3 liked a model 7 months ago
ArtusDev/Qwen_Qwen3-Coder-30B-A3B-Instruct-EXL3 liked a model 7 months ago
bartowski/Qwen2.5-Coder-32B-Instruct-exl2Organizations
None yet
liked a
model 21 days ago
liked 4
models 7 months ago
ArtusDev/Qwen_Qwen3-Coder-30B-A3B-Instruct-EXL3
Updated • 52 • 7
bartowski/Qwen2.5-Coder-32B-Instruct-exl2
Text Generation • Updated • 23 • 13
unsloth/DeepSeek-R1-0528-Qwen3-8B-GGUF
Text Generation • 8B • Updated • 51.8k • 436
unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF
Text Generation • 31B • Updated • 1.68M • 844
upvoted 2
collections 7 months ago
reacted to merterbak's post with 🔥 about 1 year ago
Post
5516
Qwen 3 models released🔥
It offers 2 MoE and 6 dense models with following parameter sizes: 0.6B, 1.7B, 4B, 8B, 14B, 30B(MoE), 32B, and 235B(MoE).
Models: Qwen/qwen3-67dd247413f0e2e4f653967f
Blog: https://qwenlm.github.io/blog/qwen3/
Demo: Qwen/Qwen3-Demo
GitHub: https://github.com/QwenLM/Qwen3
✅ Pre-trained 119 languages(36 trillion tokens) and dialects with strong translation and instruction following abilities. (Qwen2.5 was pre-trained on 18 trillion tokens.)
✅Qwen3 dense models match the performance of larger Qwen2.5 models. For example, Qwen3-1.7B/4B/8B/14B/32B perform like Qwen2.5-3B/7B/14B/32B/72B.
✅ Three stage done while pretraining:
• Stage 1: General language learning and knowledge building.
• Stage 2: Reasoning boost with STEM, coding, and logic skills.
• Stage 3: Long context training
✅ It supports MCP in the model
✅ Strong agent skills
✅ Supports seamless between thinking mode (for hard tasks like math and coding) and non-thinking mode (for fast chatting) inside chat template.
✅ Better human alignment for creative writing, roleplay, multi-turn conversations, and following detailed instructions.
It offers 2 MoE and 6 dense models with following parameter sizes: 0.6B, 1.7B, 4B, 8B, 14B, 30B(MoE), 32B, and 235B(MoE).
Models: Qwen/qwen3-67dd247413f0e2e4f653967f
Blog: https://qwenlm.github.io/blog/qwen3/
Demo: Qwen/Qwen3-Demo
GitHub: https://github.com/QwenLM/Qwen3
✅ Pre-trained 119 languages(36 trillion tokens) and dialects with strong translation and instruction following abilities. (Qwen2.5 was pre-trained on 18 trillion tokens.)
✅Qwen3 dense models match the performance of larger Qwen2.5 models. For example, Qwen3-1.7B/4B/8B/14B/32B perform like Qwen2.5-3B/7B/14B/32B/72B.
✅ Three stage done while pretraining:
• Stage 1: General language learning and knowledge building.
• Stage 2: Reasoning boost with STEM, coding, and logic skills.
• Stage 3: Long context training
✅ It supports MCP in the model
✅ Strong agent skills
✅ Supports seamless between thinking mode (for hard tasks like math and coding) and non-thinking mode (for fast chatting) inside chat template.
✅ Better human alignment for creative writing, roleplay, multi-turn conversations, and following detailed instructions.