ZeroXClem/Qwen3-4B-Hermes-Axion-Pro

Full fine-tune vs Qwen/Qwen3-4B-Thinking-2507. Fine-tune: 100% of the weights changed, across 36 of 36 layers.

Tensors changed: Norms 111/145, MLP 108/108, Attention 144/144

2 derivatives checked

• mradermacher/Qwen3-4B-Hermes-Axion-Pro-i1-GGUF: Quantization
• mradermacher/Qwen3-4B-Hermes-Axion-Pro-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse