trollek/ImagePromptHelper-danube3-500M

Full fine-tune vs h2oai/h2o-danube3-500m-base. Fine-tune: 100% of the weights changed, across 16 of 16 layers.

Tensors changed: Norms 12/33, MLP 48/48, Attention 64/64

2 derivatives checked

• mradermacher/ImagePromptHelper-danube3-500M-GGUF: Same weights, converted
• trollek/ImagePromptHelper-danube3-500M-GGUF: Same weights, converted

On the Hub · Downloads on Model Pulse