mradermacher/OceanGPT-o-7B-GGUF
LoRA-style (fewer layers) vs zjunlp/OceanGPT-o-7B. mradermacher/OceanGPT-o-7B-GGUF matches part of zjunlp/OceanGPT-o-7B (339 of its 729 tensors; the rest of zjunlp/OceanGPT-o-7B has no counterpart). On the shared part: mradermacher/OceanGPT-o-7B-GGUF is a GGUF (F16) of different weights. Looks like a merged LoRA (or a fine-tune with frozen norms): only the weight matrices of down_proj, gate_proj, k_proj, q_proj, up_proj, v_proj changed, in 28 of 28 layers; norms, biases and embeddings are untouched.
Tensors changed: Embeddings 0/1, Norms 0/57, MLP 84/84, Attention 84/196, lm_head 0/1