-
## Problem
- Model publishers may not always assign the right Model Architecture tags
- We want to detect when a model is a [TensorRT Engine](https://github.com/triton-inference-server/tensorrtllm_b…
-
#3219 introduced inference of the target versions for black from the `requires-python` project metadata in `pyproject.toml`. However, it was not implemented correctly.
The rules for the comparison …
-
Consider the following program:
```dart
void f(List xs, X Function(X) g) {}
void main() => f([], (int i) => i);
```
Both the analyzer and the CFE reject this program, reporting that "`X` coul…
-
#42 completed; now the checks need to be written
Relevant file is v7_pickle_web_interface/flask/validate_inference_rules_sympy.py
-
Allocation order inference pass has mutated allocation domain on output tensor in matmul tests. I accidentally missed the broken CI since two propagation rules going in in parallel weren't combined to…
-
possible outputs that would be nice are:
- human readable text (in the style of [EasyProve](http://easyprove.ii.uni.wroc.pl/))
- proof trees (using LaTeX and inference rules)
- Isabelle theories
- Coq…
-
The dependency inference rules need refactoring since there is duplication currently between the rule applied to `go_package` targets and the rule applied to `_go_ext_mod_package` targets (because `_g…
tdyas updated
2 years ago
-
Applies to both inference rules and latex expressions
The autocomplete feature is useful for typing in CLI; not clear how to enable for web interface.
-
**Description**
Currently, if both containers for model analyzer and Triton inference server are being deployed, so while collecting data from respective sources from their metrics endpoint port, Thi…
-
gomlx seems to call deprecated XLA Client APIs (e.g. https://github.com/gomlx/gomlx/blob/main/c/gomlx/computation.cpp#L743), but it should be using new PjRT ones.
Lots of those API calls seem auto…