Word2Li/Electricity-Meter-OCR-7B
Partial fine-tune vs Qwen/Qwen2.5-VL-7B-Instruct. Partial change: 85% of the weights changed (attention, lm_head, mlp, norms), in 28 of 32 layers.
Tensors changed: Embeddings 0/2, Norms 54/122, MLP 84/280, Attention 196/324, lm_head 1/1
1 derivatives checked
- mradermacher/Electricity-Meter-OCR-7B-GGUF: Same weights, converted (fewer layers)