google/gemma-3-4b-it-qat-int4-unquantized
Full fine-tune vs google/gemma-3-4b-it. Fine-tune: 90% of the weights changed, across 34 of 34 layers.
Tensors changed: Embeddings 1/4, Norms 206/316, MLP 102/210, Attention 136/352, other 1/1
2 derivatives checked
- DevQuasar-12/google.gemma-3-4b-it-qat-int4-unquantized-GGUF: Same weights, converted (fewer layers)
- unsloth/gemma-3-4b-it-qat-int4-unsloth-bnb-4bit: Quantization
1 other repos hold byte-for-byte identical weight files.