divelab/relay-fastdllm-v2-c40m60-distilled-step200
Full fine-tune vs Efficient-Large-Model/Fast_dLLM_v2_1.5B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 44/57, MLP 84/84, Attention 195/196
Full fine-tune vs Efficient-Large-Model/Fast_dLLM_v2_1.5B. Fine-tune: 100% of the weights changed, across 28 of 28 layers.
Tensors changed: Embeddings 1/1, Norms 44/57, MLP 84/84, Attention 195/196