oddadmix/MasriSwitch-Gemma3n-Transcriber-v1
Partial fine-tune vs unsloth/gemma-3n-E4B-it. Partial change: 52% of the weights changed (attention, embeddings, mlp, other), in 35 of 35 layers.
Tensors changed: Attention 188/363, Embeddings 1/25, Norms 0/489, other 24/687, MLP 105/112
2 derivatives checked
- mradermacher/MasriSwitch-Gemma3n-Transcriber-v1-i1-GGUF: Quantization (fewer layers)
- mradermacher/MasriSwitch-Gemma3n-Transcriber-v1-GGUF: Same weights, converted (fewer layers)