mradermacher/UnifiedReward-2.0-qwen-3b-GGUF
Full fine-tune (fewer layers) vs CodeGoat24/UnifiedReward-2.0-qwen-3b. mradermacher/UnifiedReward-2.0-qwen-3b-GGUF matches part of CodeGoat24/UnifiedReward-2.0-qwen-3b (434 of its 824 tensors; the rest of CodeGoat24/UnifiedReward-2.0-qwen-3b has no counterpart). On the shared part: mradermacher/UnifiedReward-2.0-qwen-3b-GGUF is a GGUF (F16) of different weights. Fine-tune: 100% of the weights changed, across 36 of 36 layers.
Tensors changed: Embeddings 1/1, Norms 71/73, MLP 108/108, Attention 252/252