|
pulsatrix 1.1.0
|
Which output both models are asked to explain. More...
#include <fine_tune_diff.hpp>
Public Attributes | |
| int64_t | position = -1 |
For outputs shaped (N, L, C), such as a language model's logits: the position, with negative values counted from the end. -1, the default, is the last. Ignored for (N, C). | |
| std::vector< int64_t > | classes |
| One class or token per sample. Empty: the fine-tuned model's top prediction, so both models explain the choice the fine-tune makes. | |
Which output both models are asked to explain.
| std::vector<int64_t> pulsatrix::DiffTarget::classes |
One class or token per sample. Empty: the fine-tuned model's top prediction, so both models explain the choice the fine-tune makes.
| int64_t pulsatrix::DiffTarget::position = -1 |
For outputs shaped (N, L, C), such as a language model's logits: the position, with negative values counted from the end. -1, the default, is the last. Ignored for (N, C).