nassimjp/Qwen3.5-2B-Base-Pashto-Plus
Same weights, converted (fewer layers) vs Qwen/Qwen3.5-2B-Base. nassimjp/Qwen3.5-2B-Base-Pashto-Plus matches part of Qwen/Qwen3.5-2B-Base (319 of its 632 tensors; the rest of Qwen/Qwen3.5-2B-Base has no counterpart). On the shared part: Same weights stored in another dtype (BF16/F32 → BF16).
Tensors changed: Attention 0/168, Norms 0/79, MLP 0/72