-
Thanks for your code first. I trained the model of top-down, the score on karpathy split is similar to the paper. However, when I want to test the model in test sever online of MSCOCO, the scores of m…
-
Reported by user:
### Describe the bug
1. Run setup_CESM_hybrid_ensemble.csh or setup_CESM_perfect_mod.csh
2. What was the expected outcome?
sets up an experiment
3. What actually happened? …
-
**Description**
low speed in large concurrent requests
concurrent requests | 1 | 50 | 100
-- | -- | -- | --
TensorRT-llm| 73.36 | 193.30 | 193.81
Vllm| 64.13| 984.55| 1246.50
value is TPS …
-
Continue work started in this PR:
https://github.com/mlcommons/modelgauge/pull/560
Goal: decouple test definitions from test operations.
This is step 2 of potentially 6. I think this will enable…
-
The experimenter can perform a number of functions, each with a variety of sub functions. We should consider breaking this out so that it can be more easily run and unittested.
We have:
1. Pipe…
-
After review, relabel to 'reviewTwo'. After second review, relabel to 'EditorsComment'.
-
We need to add a way to easily save the base models (with cvpreds) to disk such that someone can load them into a future H2O session and use the h2o.stack() function. This is similar to the h2o.save_…
-
Hello Wang,
There is an error when I set batch_size to more than 1.
Firstly, some dismatch issue occrured at DenseNet forward at models.py row 151. But I fixed it by set `out = out.view(out.size(…
-
Instead of optimizing the control trajectory based on a deterministic model, it is possible to optimize them based on an ensemble of models, i.e. a probability distribution of models, each of which is…
-
Hello,
I have an ensemble model that returns weighted average of results from elasticnet/ridge/xgboost/randomforest. I understand I can use linear explainer for elasticnet and ridge, and tree expla…