kjhq/ERNIE-4.5-0.3B-Distilled

Full fine-tune vs baidu/ERNIE-4.5-0.3B-Base-PT. Fine-tune: 100% of the weights changed, across 18 of 18 layers.

Tensors changed: Embeddings 1/1, Norms 2/37, MLP 54/54, Attention 72/72

1 derivatives checked

• mradermacher/ERNIE-4.5-0.3B-Distilled-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse