Mungert/DeepCoder-14B-Preview-GGUF
Partial fine-tune vs deepseek-ai/DeepSeek-R1-Distill-Qwen-14B. Mungert/DeepCoder-14B-Preview-GGUF is a GGUF (Q8_0) of different weights. Partial change: 89% of the weights changed (attention, mlp, norms), in 48 of 48 layers.
Tensors changed: Embeddings 0/1, Attention 336/336, MLP 144/144, Norms 92/97, lm_head 0/1