Community User
Displaying 1 to 14 of 14 repositories
Qwen3.5-122B-A10B INT4 (thinking+8192) for the K-AI leaderboard. served-model-name JGOS-125B-Q4.
2m
10K+
JGOS-31B-Citizen — K-AI Leaderboard #1 Korean LLM by VIDRAFT
3m
810
Darwin-36B-Opus, VKAE-accelerated: 280.8 tok/s single-GPU (11.2x), OpenAI-compatible.
3m
288
VKAE-accelerated Qwen3.5-35B-A3B - 601 tok/s single-GPU, OpenAI-compatible serving in one image.
3m
1.6K
VIDRAFT 397B Korean LLM. HF: https://huggingface.co/FINAL-Bench/JGOS-398B-fp8
4m
120