Cas-Warehouse/Llama-3-Depressed-Therapist-8B
LoRA-style vs Cas-Warehouse/Llama-3-SOVL-MopeyMule-Blackroot-8B. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, o_proj, q_proj, v_proj changed, in 32 of 32 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/65, MLP 4/96, Attention 57/128
4 derivatives checked
- mradermacher/Llama-3-Deppressed-Therapist-8B-i1-GGUF: Quantization
- mradermacher/Llama-3-Depressed-Therapist-8B-GGUF: Same weights, converted
- mradermacher/Llama-3-Deppressed-Therapist-8B-GGUF: Same weights, converted
- Cas-Archive/L3-Umbral-Mind-RP-v0.6.2-8B: Full fine-tune