issues
search
neuralmagic
/
deepsparse
Sparsity-aware deep learning inference runtime for CPUs
https://neuralmagic.com/deepsparse/
Other
2.97k
stars
171
forks
source link
[Text Generation][Pipeline Refactor] Add input tokens to output for perplexity
#1474
Closed
dsikka
closed
8 months ago
dsikka
commented
8 months ago
Summary
This PR was merged while we were working on v2. Involves adding input_tokens to the output for perplexity
Updates to no longer skip the test for this
Summary