perplexity-ai/browsesafe

Full fine-tune vs Qwen/Qwen3-30B-A3B-Instruct-2507. Fine-tune: 100% of the weights changed, across 48 of 48 layers.

Tensors changed: Embeddings 1/1, Norms 191/193, MLP 2613/2613, Attention 192/192, lm_head 1/1

1 derivatives checked

• DevQuasar/perplexity-ai.browsesafe-GGUF: Same weights, converted

1 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse