fdtn-ai/Foundation-Sec-8B
Full fine-tune vs meta-llama/Llama-3.1-8B. Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Norms 65/65, MLP 96/96, Attention 128/128
10 derivatives checked
- fdtn-ai/Foundation-Sec-8B-Instruct: Full fine-tune
- fdtn-ai/Foundation-Sec-1.1-8B-Instruct: Full fine-tune
- fdtn-ai/Foundation-Sec-8B-Reasoning: Full fine-tune
- Mungert/Foundation-Sec-8B-Instruct-GGUF: LoRA-style
- mradermacher/Foundation-Sec-8B-i1-GGUF: Quantization
- fdtn-ai/Foundation-Sec-8B-Q4_K_M-GGUF: Quantization
- 2p8xx/Foundation-Sec-8B-Instruct-GGUF: LoRA-style
- mradermacher/Foundation-Sec-8B-GGUF: Same weights, converted
- DevQuasar/fdtn-ai.Foundation-Sec-8B-GGUF: Same weights, converted
- fdtn-ai/Foundation-Sec-8B-Q8_0-GGUF: Quantization
3 other repos hold byte-for-byte identical weight files.