sainikhiljuluri2015/Apriel-1.5-15b-Thinker-CYBERSEC-MERGED
LoRA-style vs ServiceNow-AI/Apriel-1.5-15b-Thinker. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, o_proj, q_proj, up_proj, v_proj changed, in 47 of 48 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/146, MLP 78/216, Attention 91/288, other 0/5, lm_head 0/1
2 derivatives checked
- mradermacher/Apriel-1.5-15b-Thinker-CYBERSEC-MERGED-i1-GGUF: Quantization (fewer layers)
- mradermacher/Apriel-1.5-15b-Thinker-CYBERSEC-MERGED-GGUF: Quantization (fewer layers)