NousResearch/Hermes-4.3-36B
Full fine-tune vs ByteDance-Seed/Seed-OSS-36B-Base. Fine-tune: 100% of the weights changed, across 64 of 64 layers.
Tensors changed: Embeddings 1/1, Norms 22/129, MLP 192/192, Attention 447/448, lm_head 1/1
10 derivatives checked
- bartowski/NousResearch_Hermes-4.3-36B-GGUF: Same weights, converted
- mradermacher/Hermes-4.3-36B-i1-GGUF: Quantization
- cyankiwi/Hermes-4.3-36B-AWQ-4bit: Quantization
- DevQuasar/NousResearch.Hermes-4.3-36B-GGUF: Same weights, converted
- Doradus-AI/Hermes-4.3-36B-FP8: Quantization
- mradermacher/Hermes-4.3-36B-GGUF: Quantization
- MaziyarPanahi/Hermes-4.3-36B-GGUF: Same weights, converted
- lukaskremla/Hermes-4.3-36B-mlx-5Bit: Quantization
- alexcovo/Hermes-4.3-36B-mlx-8Bit: Quantization
- alexcovo/Hermes-4.3-36B-mlx-4Bit: Quantization
7 other repos hold byte-for-byte identical weight files.