moonshotai/Kimi-VL-A3B-Thinking-2506
Full fine-tune vs moonshotai/Kimi-VL-A3B-Instruct. Fine-tune: 100% of the weights changed, across 27 of 27 layers.
Tensors changed: Embeddings 4/4, Norms 194/194, MLP 2553/2554, Attention 162/189, other 58/58, lm_head 1/1
10 derivatives checked
- ggml-org/Kimi-VL-A3B-Thinking-2506-GGUF: Same weights, converted
- mradermacher/Kimi-VL-A3B-Thinking-2506-GGUF: Quantization
- mradermacher/Kimi-VL-A3B-Thinking-2506-i1-GGUF: Quantization
- SoybeanMilk/Kimi-VL-A3B-Thinking-2506-BNB-4bit: Quantization
- richardyoung/Kimi-VL-A3B-Thinking-GGUF: Quantization
- DevQuasar/moonshotai.Kimi-VL-A3B-Thinking-2506-GGUF: Quantization
- blascotobasco/Kimi-VL-A3B-Thinking-2506-APEX-GGUF: Quantization
- AbteeXAILab/lumynax-multimodal-kimi-vl-a3b-thinking: Quantization
- noctrex/Kimi-VL-A3B-Thinking-2506-MXFP4_MOE-GGUF: Quantization
- ssweens/Kimi-VL-A3B-Thinking-2506-GGUF: Same weights, converted