Nanbeige/Nanbeige4.2-3B
Full fine-tune vs Nanbeige/Nanbeige4.2-3B-Base. Fine-tune: 100% of the weights changed, across 22 of 22 layers.
Tensors changed: lm_head 1/1, Embeddings 1/1, Norms 45/45, MLP 66/66, Attention 88/88
27 derivatives checked
- bartowski/Nanbeige_Nanbeige4.2-3B-GGUF: Same weights, converted
- owao/Nanbeige4.2-3B-GGUF: Quantization
- gkraker04/Nanbeige4.2-3B-GGUF: Same weights, converted
- Tdamre/Nanbeige4.2-3B-GGUF: Quantization
- NANI-Nithin/Nanbeige4.2-3B-GGUF: Same weights, converted
- tinyopsec/Nanbeige4.2-3B-GGUF: Same weights, converted
- webmp3/Sakura-Nanbeige4.2-3B-GGUF: Quantization
- AnkitAI/Parable-Nanbeige4.2-3B-Claude-Fable-5-GGUF: LoRA-style
- mradermacher/Nanbeige4.2-3B-i1-GGUF: Quantization
- iamimmanuelraj/Nanbeige4.2-3B-GGUF: Same weights, converted
- mradermacher/Nanbeige4.2-3B-GGUF: Same weights, converted
- Nanbeige/Nanbeige4.2-3B-FP8: Quantization
- nicolasembleton/Nanbeige4.2-3B-GGUF: Same weights, converted
- mlx-community/Nanbeige4.2-3B-OptiQ-4bit: Quantization
- Abiray/Nanbeige4.2-3B-GGUF: Quantization
- Soulfate24/Nanbeige4.2-3B-DSpark-Paretrix: Full fine-tune
- MercuriusDream/Nanbeige4.2-3B-mlx-8bit: Quantization
- AnkitAI/Parable-Nanbeige4.2-3B-Claude-Fable-5: LoRA-style
- MercuriusDream/Nanbeige4.2-3B-mlx-4bit: Quantization
- ProCreations/grug-3b: LoRA-style
- Andgihat/Nanbeige4.2-3B-GGUF: Quantization
- Nanbeige/Nanbeige4.2-3B-GPTQ-Int8: Quantization
- OsaurusAI/Nanbeige4.2-3B-JANG_6M: Quantization
- DevQuasar/Nanbeige.Nanbeige4.2-3B-GGUF: Quantization
- Irfanuruchi/Nanbeige4.2-3B-CE-v1.0: LoRA-style
- WaveCut/Nanbeige4.2-3B-heretic: Abliterated
- MuXodious/Nanbeige4.2-3B-ARA-heresy: Abliterated
1 other repos hold byte-for-byte identical weight files.