mradermacher/TinyStories-LLaMA2-20M-256h-4l-GQA-GGUF
LoRA-style vs chencindy/TinyStories-LLaMA2-20M-256h-4l-GQA. mradermacher/TinyStories-LLaMA2-20M-256h-4l-GQA-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 k_proj, q_proj changed, in 4 of 4 layers; norms, biases and embeddings are untouched.
Tensors changed: lm_head 0/1, Embeddings 0/1, Norms 0/9, MLP 0/12, Attention 8/16