EVA-UNIT-01/EVA-Qwen2.5-72B-v0.0
Partial fine-tune vs Qwen/Qwen2.5-72B. Partial change: 50% of the weights changed (attention, lm_head, mlp), in 80 of 80 layers.
Tensors changed: Embeddings 0/1, MLP 120/240, Attention 280/560, Norms 0/161, lm_head 1/1
5 derivatives checked
- bartowski/EVA-Qwen2.5-72B-v0.0-GGUF: Quantization
- mradermacher/EVA-Qwen2.5-72B-v0.0-i1-GGUF: Quantization
- mradermacher/EVA-Qwen2.5-72B-v0.0-GGUF: Quantization
- mlx-community/EVA-Qwen2.5-72B-v0.0: Same weights, converted
- KaraKaraWarehouse/LLENN-v0.69420-Qwen2.5-72b: Full fine-tune