mlx-community/deepseek-r1-distill-qwen-1.5b

Same weights, converted vs deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B. Same weights stored in another dtype (BF16 → F16).

Tensors changed: Embeddings 0/1, Attention 0/196, MLP 0/84, Norms 0/57, lm_head 0/1

3 other repos hold byte-for-byte identical weight files.

On the Hub · Downloads on Model Pulse