artificialguybr/LLAMA-3.2-1B-OpenHermes2.5
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 9/33, MLP 48/48, Attention 64/64
8 derivatives checked
- artificialguybr/LLAMA-3.2-1B-OpenHermes2.5-gguf: Same weights, converted
- mradermacher/LLAMA-3.2-1B-OpenHermes2.5-GGUF: Same weights, converted
- WithinUsAI/Llama3.2-Hermes.Dolphin.Coder-1B: Full fine-tune
- Nexesenex/Llama_3.2_1b_Syneridol_0.2: Full fine-tune
- Nexesenex/Llama_3.2_1b_Synopsys_0.1: Full fine-tune
- Nexesenex/Llama_3.2_1b_Synopsys_0.11: Full fine-tune
- Nexesenex/Llama_3.2_1b_Synerka_0.0: Full fine-tune
- Nexesenex/Llama_3.2_1b_Odyssea_Escalation_0.2: Full fine-tune