-
lightblue/DeepSeek-R1-Distill-Qwen-1.5B-Multilingual
2B • Updated • 13 • 24 -
lightblue/DeepSeek-R1-Distill-Qwen-7B-Multilingual
8B • Updated • 8 • 22 -
lightblue/DeepSeek-R1-Distill-Qwen-14B-Multilingual
15B • Updated • 68 • 13 -
lightblue/DeepSeek-R1-Distill-Qwen-7B-Japanese
Text Generation • 8B • Updated • 13 • 33
AI & ML interests
None defined yet.
Multipurpose RAG models for many languages
-
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-full
Text Generation • 8B • Updated • 8.18k • 2 -
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top75
Text Generation • 8B • Updated • 8.18k • 4 -
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half
Text Generation • 8B • Updated • 8.16k • 16 -
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top25
Text Generation • 8B • Updated • 8.82k • 3
-
lightblue/DeepSeek-R1-Distill-Qwen-1.5B-Multilingual
2B • Updated • 13 • 24 -
lightblue/DeepSeek-R1-Distill-Qwen-7B-Multilingual
8B • Updated • 8 • 22 -
lightblue/DeepSeek-R1-Distill-Qwen-14B-Multilingual
15B • Updated • 68 • 13 -
lightblue/DeepSeek-R1-Distill-Qwen-7B-Japanese
Text Generation • 8B • Updated • 13 • 33
The models trained under our Karasu and Qarasu project
Multipurpose RAG models for many languages
Our latest fine-tuned models
-
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-full
Text Generation • 8B • Updated • 8.18k • 2 -
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top75
Text Generation • 8B • Updated • 8.18k • 4 -
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-half
Text Generation • 8B • Updated • 8.16k • 16 -
lightblue/suzume-llama-3-8B-multilingual-orpo-borda-top25
Text Generation • 8B • Updated • 8.82k • 3