swiss-ai/Apertus-8B-Instruct-2509
Full fine-tune vs swiss-ai/Apertus-8B-2509. Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 1/1, Norms 117/129, MLP 84/192, Attention 128/128, lm_head 1/1
24 derivatives checked
- unsloth/Apertus-8B-Instruct-2509-GGUF: Same weights, converted
- EPFLiGHT/Apertus-8B-MeditronFO: Full fine-tune
- bartowski/swiss-ai_Apertus-8B-Instruct-2509-GGUF: Same weights, converted
- mradermacher/Apertus-8B-Instruct-2509-i1-GGUF: Quantization
- giladgd/Apertus-8B-Instruct-2509-GGUF: Same weights, converted
- RedHatAI/Apertus-8B-Instruct-2509-FP8-dynamic: Quantization
- redponike/Apertus-8B-Instruct-2509-GGUF: Quantization
- mradermacher/Apertus-8B-Instruct-2509-GGUF: Same weights, converted
- saidutta69/Apertus-8B-Instruct-heretic: Abliterated
- mlx-community/Apertus-8B-Instruct-2509-bf16: Same weights, repacked
- gaparecido/apertus-8b-financial-reasoner-v1: LoRA-style
- thomaskiefer/EAGLE3-Apertus-8B-Instruct-2509: Different architecture
- aisingapore/Apertus-SEA-LION-v4-8B-IT: Full fine-tune
- MaziyarPanahi/Apertus-8B-Instruct-2509-GGUF: Same weights, converted
- aisingapore/Apertus-SEA-LION-v4-8B-IT-GGUF: Full fine-tune
- anicka/karma-electric-apertus-8b: LoRA-style
- mlx-community/Apertus-8B-Instruct-2509-4bit: Quantization
- DevQuasar/swiss-ai.Apertus-8B-Instruct-2509-GGUF: Quantization
- anicka/ke-v9-8b: LoRA-style
- loleg/Apertus-8B-Instruct-2509-mlx: Same weights, repacked
- gaparecido/apertus-8b-financial-reasoner-v2: Adapter
- Colby/apertus-8b-coding-gguf: LoRA-style
- mlx-community/Apertus-8B-Instruct-2509-8bit: Quantization
- tartuNLP/Apertus-EstLLM-8B-Instruct-0326: Full fine-tune
3 other repos hold byte-for-byte identical weight files.