Hugging Face Forums
Carrying Gradients Through Generate
Research
patrickvonplaten
November 2, 2020, 12:55pm
5
This should also be interesting:
Big `generate()` refactor
show post in topic
Related topics
Topic
Replies
Views
Activity
A potential in-place operation that caused an RuntimeError
🤗Transformers
1
2318
January 19, 2021
T5 user defined loss function
Beginners
11
4812
September 23, 2020
T5 forward pass versus generate, latter outputs non-sense
Beginners
8
2918
March 25, 2021
Disparity between output from `forward` and `generate` for greedy search (using Whisper)
🤗Transformers
3
1378
August 11, 2024
Language model gradients sensitive to target value/length
Research
0
347
June 16, 2023