update end-to-end example to use systems api#680
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
Documentation previewhttps://nvidia-merlin.github.io/Transformers4Rec/review/pr-680 |
| @@ -112,10 +112,10 @@ | |||
| "name": "stderr", | |||
There was a problem hiding this comment.
Line #2. table = TensorTable.from_df(df.iloc[:100])
can you please add a new cell and add these two lines ( we need my PR to be merged first for the top_k part works ):
Let's now add a top_k parameter to model so that we can return the top k item ids with the highest scores when we serve the model on Triton Inference Server.
model.top_k = 28
model.eval()
Reply via ReviewNB
There was a problem hiding this comment.
Line #2. table = TensorTable.from_df(df.iloc[:100])
can you please add a new cell and add these two lines ( we need my PR to be merged first for the top_k part works ):
Let's now add a top_k parameter to model so that we can return the top k item ids with the highest scores when we serve the model on Triton Inference Server.
model.top_k = 28
model.eval()Reply via ReviewNB
you can set top_k 10 or something else, if you want. I set it to 28 just as an example.
ad9f35c to
d9db33d
Compare
774430c to
62e1fbb
Compare
resolves #676