Unverified50% confidenceFactExact time
The asmsense model uses Qwen3-8B as the base model and is fine-tuned via LoRA.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/18/2026
First Seen
Valid until: 1/16/2027
Sources
asmsense: 228K x86-64 Function Dataset Tackles Reverse Engineering Annotation Challenges
redditr/MachineLearning10/18/2026
Related Entities
Related Claims
UnverifiedThis version supports exporting Clef models based on the Qwen3.5 backbone, as well as Laya models, to GGUF format.71% similarUnverifiedAll of these Qwen3 uncensored models provide GGUF format support and retain native MTP (Multi-Token Prediction) capability.68% similarVerifiedOllama natively supports the GGUF format, and most models have been pre-quantized to specifications such as Q4_K_M or Q8_0.67% similarUnverifiedQAT versions of Gemma models are directly compatible with local large model tools such as Ollama and LM Studio, and can be used immediately after downloading.66% similarUnverifiedThe model trained a rank-256 LoRA adapter on top of the base model huihui-ai/Huihui-Qwen3.8-27B-abliterated.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1061121API
curl https://kongchang.com/api/v1/knowledge/claims/1061121MCP
get_claim(id=1061121)