fastino/GLiNER2.5-Decide
Partial fine-tune vs fastino/gliner2-large-v1. Partial change: 90% of the weights changed (attention, embeddings, norms, other), in 24 of 24 layers.
Tensors changed: other 100/116, Embeddings 4/13, Norms 98/98, Attention 192/192
4 derivatives checked
- aufklarer/GLiNER2.5-Decide-340M-MLX-8bit: Quantization
- antflydb/gliner2.5-decide-gguf: Quantization
- aufklarer/GLiNER2.5-Decide-340M-MLX-fp16: Different architecture
- aufklarer/GLiNER2.5-Decide-340M-MLX: Different architecture
3 other repos hold byte-for-byte identical weight files.