zjunlp/OceanGPT-basic-7B-v0.3
LoRA-style vs Qwen/Qwen2.5-7B-Instruct. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of q_proj, v_proj changed, in 28 of 28 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/57, MLP 0/84, Attention 56/196, lm_head 0/1
4 derivatives checked
- mradermacher/OceanGPT-basic-7B-i1-GGUF: Quantization
- mradermacher/OceanGPT-basic-7B-v0.3-i1-GGUF: Quantization
- mradermacher/OceanGPT-basic-7B-v0.3-GGUF: Same weights, converted
- mradermacher/OceanGPT-basic-7B-GGUF: Same weights, converted