bunnycore/Qwen2.5-7B-RRP-1M

Full fine-tune vs Qwen/Qwen2.5-7B-Instruct-1M. Fine-tune: 100% of the weights changed, across 28 of 28 layers.

Tensors changed: Norms 6/57, MLP 84/84, Attention 115/196

3 derivatives checked

• mradermacher/Qwen2.5-7B-RRP-1M-i1-GGUF: Quantization
• mradermacher/Qwen2.5-7B-RRP-1M-GGUF: Same weights, converted
• Sakalti/SJT-7B-V1.1: Full fine-tune

On the Hub · Downloads on Model Pulse