yasserrmd/Fara-TARS-7B

Full fine-tune vs ByteDance-Seed/UI-TARS-1.5-7B. Fine-tune: 100% of the weights changed, across 32 of 32 layers.

Tensors changed: Embeddings 2/2, Attention 324/324, MLP 280/280, Norms 75/122, lm_head 1/1

2 derivatives checked

• mradermacher/Fara-TARS-7B-i1-GGUF: Quantization (fewer layers)
• mradermacher/Fara-TARS-7B-GGUF: Same weights, converted (fewer layers)

On the Hub · Downloads on Model Pulse