LLM-OS-Models/Fabliq-8B-Agent-FromBase

Full fine-tune vs LiquidAI/LFM2.5-8B-A1B. Fine-tune: 97% of the weights changed, across 24 of 24 layers.

Tensors changed: MLP 2140/2162, Embeddings 0/2, other 54/54, Norms 9/60, Attention 24/24

3 derivatives checked

• mradermacher/Fabliq-8B-Agent-FromBase-i1-GGUF: Quantization
• mradermacher/Fabliq-8B-Agent-FromBase-GGUF: Same weights, converted
• LLM-OS-Models/Fabliq-8B-Agent-FromBase-Reasoning: Full fine-tune

On the Hub · Downloads on Model Pulse