Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash
LoRA-style vs unsloth/Qwen3.5-9B. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, out_proj, q_proj, up_proj, v_proj changed, in 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/5, MLP 96/211, Attention 56/336, other 0/1, Norms 0/221
9 derivatives checked
- mradermacher/Qwen3.5-9B-DeepSeek-V4-Flash-i1-GGUF: Quantization
- mradermacher/Qwen3.5-9B-DeepSeek-V4-Flash-GGUF: Same weights, converted
- Jackrong/MLX-Qwen3.5-9B-DeepSeek-V4-Flash-4bit: Quantization
- Jackrong/Qwen3.5-9B-DeepSeek-V4-Flash-MTP-GGUF: Quantization
- Arien0/Qwen3.5-9B-DeepSeek-V4-Flash-AWQ-BF16-INT8-UC-MTP: Quantization
- Sohailhosseini/Qwen3.5-9B-DeepSeek-V4-Flash-AWQ-W4A16: Quantization
- TheWirelessPhoenix/Qwen3.5-9B-DeepSeek-V4-Flash-4bit-vmlx: Quantization
- tamerbaydag/Qwen3.5-9B-DeepSeek-V4-Flash-Q4_K_M-GGUF: Quantization
- jorge-erdb/Qwen3.5-9B-DeepSeek-V4-Flash-4bit-GGUF: Quantization