chakra-labs/GLADOS-1

Partial fine-tune vs ByteDance-Seed/UI-TARS-7B-SFT. Partial change: 0% of the weights changed (attention, norms), in 11 of 32 layers.

Tensors changed: Embeddings 0/2, Attention 5/324, MLP 0/216, Norms 7/187, lm_head 0/1

2 derivatives checked

• mradermacher/GLADOS-1-i1-GGUF: Quantization (fewer layers)
• mradermacher/GLADOS-1-GGUF: Same weights, converted (fewer layers)

On the Hub · Downloads on Model Pulse