TokenRhythm/NeoHorse-1-9B
Partial fine-tune (fewer layers) vs Qwen/Qwen3.5-9B. TokenRhythm/NeoHorse-1-9B matches part of Qwen/Qwen3.5-9B (427 of its 775 tensors; the rest of Qwen/Qwen3.5-9B has no counterpart). On the shared part: Partial change: 89% of the weights changed (attention, lm_head, mlp, norms), in 32 of 32 layers.
Tensors changed: lm_head 1/1, Embeddings 0/1, MLP 96/96, Attention 182/224, Norms 70/105
19 derivatives checked
- TokenRhythm/NeoHorse-1-9B-GGUF: Same weights, converted
- ngquocvinh/NeoHorse-1-9B-GGUF: Quantization
- wepiqx/NeoHorse-1-9B-MERNIK-GGUF: Quantization
- mradermacher/NeoHorse-1-9B-i1-GGUF: Quantization
- mradermacher/NeoHorse-1-9B-GGUF: Same weights, converted
- prithivMLmods/NeoHorse-1-9B-GGUF: Same weights, converted
- TokenRhythm/NeoHorse-1-9B-MLX-8bit: Quantization
- TokenRhythm/NeoHorse-1-9B-MLX-4bit: Quantization
- Indexnusrefather/Palette-RP-9B-2609-v0.1: LoRA-style
- Jeethu/NeoHorse-1-9B-PARO: Quantization
- TokenRhythm/NeoHorse-1-9B-MLX: Partial fine-tune
- OliviaRossi/MiMo-NeoHorse-9B-AGSI: Full fine-tune (fewer layers)
- SirSahOl/NeoHorse-1-9B-chat-mlx-16bit: Partial fine-tune
- temaq-org/Tema_Q-X6-9B-Thinking: LoRA-style
- rapid-mlx/NeoHorse-1-9B-MLX-4bit: Quantization
- OliviaRossi/MiMo-NeoHorse-9B-AGSI-GGUF: Full fine-tune
- Takenoko12345678/NeoHorse-1-9B-Math-SFT: LoRA-style
- Cenedril/Gullfaxi-9B-Q4_K_M-GGUF: Quantization
- cbert33/NeoHorse-1-9B-FP8-Calibrated: Quantization
1 other repos hold byte-for-byte identical weight files.