OzLabs/Caspi-1.7B

Full fine-tune vs Qwen/Qwen3-ASR-1.7B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.

Tensors changed: other 11/11, MLP 180/180, Norms 163/211, Attention 304/304, Embeddings 1/1

1 derivatives checked

• mradermacher/Caspi-1.7B-GGUF: Same weights, converted (fewer layers)

On the Hub · Downloads on Model Pulse