Unverified50% confidenceFactExact time
The working solution relies on a combination: DeepSeek provides the weights, coolbho3k does the 3-bit quantization, turboderp wrote the EXL3 format and the ExLlamaV3 inference library, Victor Cruz wrote the vllm-exl3 bridge layer, Diffbot designed the per-expert split scheme, startup flow and kernels, and InferenceKey did the integration and measurement.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/16/2026
First Seen
Valid until: 1/14/2027
Sources
Related Entities
Related Claims
UnverifiedEXL3 is a 3-bit quantization format introduced by ExLlamaV3 and is the successor to EXL2; ExLlamaV3 is an inference library maintained by turboderp; vllm-exl3 is a bridge layer written by Victor Cruz that connects ExLlamaV3's quantization kernels into vLLM's scheduling and PagedAttention system.76% similarUnverifiedExLlamaV3 是专为消费级 NVIDIA GPU 上高效运行 GGUF/EXL2 等量化格式大语言模型而设计的推理引擎65% similarUnverifiedAdaptiveSpec supports three mainstream target models: DeepSeek-R1-Distill-Llama-8B, Llama-3.1-8B-Instruct, and Qwen3-8B.64% similarUnverifiedDeepSeek V3和R1采用MoE（混合专家）架构，使Mac Studio或DGX Spark这类设备能以可用速度运行大模型63% similarUnverified围绕开放权重模型涌现出llama.cpp、vLLM、Ollama、Text Generation Inference等推理框架，其中llama.cpp支持CPU和Apple Silicon推理，vLLM是高吞吐GPU推理引擎62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1044817API
curl https://kongchang.com/api/v1/knowledge/claims/1044817MCP
get_claim(id=1044817)