mradermacher/VGPO-RL-32B-GGUF

Quantization (fewer layers) vs MuMing0102/VGPO-RL-32B. mradermacher/VGPO-RL-32B-GGUF matches part of MuMing0102/VGPO-RL-32B (771 of its 1161 tensors; the rest of MuMing0102/VGPO-RL-32B has no counterpart). On the shared part: mradermacher/VGPO-RL-32B-GGUF is a GGUF quantization (Q8_0) of the same weights: 321 tensors match exactly and 450 within quantization error.

Tensors changed: MLP 0/192, Attention 0/448, Norms 0/129, lm_head 0/1, Embeddings 0/1

On the Hub · Downloads on Model Pulse