Drop-Upcycling
updated
9B • Updated • 96
9B • Updated • 99
19B • Updated • 95
9B • Updated • 102
9B • Updated • 95
0.4B • Updated • 95
0.4B • Updated • 94
0.4B • Updated • 93
0.4B • Updated • 97
9B • Updated • 87
19B • Updated • 104
0.4B • Updated • 97
9B • Updated • 90
0.4B • Updated • 97
• 1
2B • Updated • 108
0.2B • Updated • 114
4B • Updated • 89
14B • Updated • 111
llm-jp/Dense-btx-code-expert-152M
0.2B • Updated • 103
• 1
llm-jp/Dense-btx-english-expert-1.5B
2B • Updated • 99
llm-jp/Dense-btx-code-expert-1.5B
2B • Updated • 105
• 1
llm-jp/Dense-btx-japanese-expert-1.5B
2B • Updated • 191
• 1
llm-jp/Dense-btx-english-expert-152M
0.2B • Updated • 100
llm-jp/Dense-btx-japanese-expert-152M
0.2B • Updated • 135
Drop-Upcycling: Training Sparse Mixture of Experts with Partial
Re-initialization
Paper
• 2502.19261
• Published • 6
Text Generation
• 73B • Updated • 119
llm-jp/llm-jp-3-8x13b-instruct3
Text Generation
• 73B • Updated • 271
• 8
llm-jp/llm-jp-3-8x1.8b-instruct3
Text Generation
• 9B • Updated • 119
• 4
Text Generation
• 9B • Updated • 253
llm-jp/llm-jp-3-8x13b-instruct2
Text Generation
• 73B • Updated • 244
llm-jp/llm-jp-3-8x1.8b-instruct2
Text Generation
• 9B • Updated • 111
llm-jp/llm-jp-3.1-8x13b-instruct4
Text Generation
• 73B • Updated • 197
• 4
Text Generation
• 73B • Updated • 123