yes, unfortunately :( did you know that the weight adapter happens on demand, on inference, in the LoRA forward pass function? leaky, leaky, leaky GGML will

yes, unfortunately :( did you know that the weight adapter happens on demand, on inference, in the LoRA forward pass function? leaky, leaky, leaky GGML will save us, surely