GenTrendGPT/OS-Test-Mark-GEN-I
Same weights, repacked vs Qwen/Qwen1.5-4B-Chat. Same weights in different files (re-saved, re-sharded or new metadata), plus 144 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/60, MLP 0/90, Attention 0/210
Same weights, repacked vs Qwen/Qwen1.5-4B-Chat. Same weights in different files (re-saved, re-sharded or new metadata), plus 144 new tensors.
Tensors changed: Embeddings 0/1, Norms 0/60, MLP 0/90, Attention 0/210