UNIVA-Bllossom/DeepSeek-llama3.1-Bllossom-8B
Full fine-tune vs deepseek-ai/DeepSeek-R1-Distill-Llama-8B. Fine-tune: 93% of the weights changed, across 32 of 32 layers.
Tensors changed: Embeddings 0/1, Attention 128/128, MLP 96/96, Norms 30/65, lm_head 1/1
4 derivatives checked
- DevQuasar-7/UNIVA-Bllossom.DeepSeek-llama3.1-Bllossom-8B-GGUF: Full fine-tune
- mradermacher/DeepSeek-llama3.1-Bllossom-8B-GGUF: Full fine-tune
- MaziyarPanahi/DeepSeek-llama3.1-Bllossom-8B-GGUF: Full fine-tune
- KYUNGYONG/DeepSeek-llama3.1-Bllossom-8B-Q4-mlx: Quantization