dphn/Dolphin3.0-Llama3.2-1B
Full fine-tune vs meta-llama/Llama-3.2-1B. Fine-tune: 100% of the weights changed, across 16 of 16 layers.
Tensors changed: Norms 33/33, MLP 48/48, Attention 64/64
13 derivatives checked
- bartowski/Dolphin3.0-Llama3.2-1B-GGUF: Same weights, converted
- mradermacher/Dolphin3.0-Llama3.2-1B-i1-GGUF: Quantization
- Novaciano/DOLPHIN3.0-L3.2-1B_RP_UNCENSORED-GGUF: Quantization
- mlx-community/dolphin3.0-llama3.2-1B-4Bit: Quantization
- DevQuasar-7/cognitivecomputations.Dolphin3.0-Llama3.2-1B-GGUF: Same weights, converted
- mradermacher/Dolphin3.0-Llama3.2-1B-GGUF: Same weights, converted
- itlwas/Dolphin3.0-Llama3.2-1B-Q4_K_M-GGUF: Quantization
- ysn-rfd/Dolphin3.0-Llama3.2-1B-GGUF: Quantization
- Nexesenex/Llama_3.2_1b_Dolto_0.1: Full fine-tune
- Nexesenex/Dolphin3.0-Llama3.1-1B-abliterated: Abliterated
- SuperPeaceBusters/Jintan-LLaMA3.2-1B-v1: LoRA-style
- SuperPeaceBusters/Menma-LLaMA3.2-1B-v1: LoRA-style
- podsni/hades-1B: LoRA-style
2 other repos hold byte-for-byte identical weight files.