mlx-community/SmolVLM2-500M-Video-Instruct-mlx
Full fine-tune (fewer layers) vs HuggingFaceTB/SmolLM2-360M-Instruct. HuggingFaceTB/SmolLM2-360M-Instruct matches part of mlx-community/SmolVLM2-500M-Video-Instruct-mlx (289 of its 489 tensors; the rest of mlx-community/SmolVLM2-500M-Video-Instruct-mlx has no counterpart). On the shared part: Fine-tune: 100% of the weights changed, across 32 of 32 layers.
Tensors changed: Norms 65/65, MLP 96/96, Attention 128/128