dinerburger/KAT-Coder-V2.5-Dev-GGUF overview

---

license: apache-2.0

base_model:

• Kwaipilot/KAT-Coder-V2.5-Dev

---

Simple, no-matrix version of KAT-Coder-V2.5-Dev with the following recipe:

QUANT="q6_k"
llama-quantize \
  --output-tensor-type Q6_K \
  --token-embedding-type bf16 \
  --tensor-type attn_qkv=bf16 \
  --tensor-type attn_v=bf16 \
  --tensor-type attn_q=bf16 \
  --tensor-type attn_k=bf16 \
  --tensor-type attn_gate=bf16 \
  --tensor-type attn_output=bf16 \
  --tensor-type ssm_ba=bf16 \
  --tensor-type ssm_beta=bf16 \
  --tensor-type ssm_alpha=bf16 \
  --tensor-type ssm_out=bf16 \
  --tensor-type ffn_down_shexp=bf16 \
  --tensor-type ffn_gate_shexp=bf16 \
  --tensor-type ffn_up_shexp=bf16 \
  Kwaipilot_KAT-Coder-V2.5-Dev-bf16-00001-of-00002.gguf \
  KAT-Coder-V2.5-Dev.${QUANT}.gguf \
  ${QUANT}