Pyinfer is a model agnostic tool for ML developers and researchers to benchmark the inference statistics for machine learning models or functions.
Installation
pip install pyinfer
Overview
Inference Report
InferenceReport is for reporting inference statistics on a single model artifact. To create a valid report simply pass it a callable model function or method, valid input(s), and either n_iterations or n_seconds to determine what interval the report uses for its run duration. Check out the docs for more information on the optional parameters that can be passed.
Multi Inference Report
MultiInferenceReport is for reporting inference statistics on a list of model artifacts. To create a valid multi report pass it a list of callable model functions or methods, a list of valid input(s), and either n_iterations or n_seconds to determine what interval the report uses for its run duration. Check out the docs for more information on the optional parameters that can be passed.
Example Outputs
Table Report
Run Plot
Stats Currently Included
- Success Rate - Number of successful inferences within a specified time range.
- Failures - Number of inferences above specified time range.
- Time Taken - Total time taken to run all inferences.
- Inference Per Second - Estimate of how many inferences per second the selected model can perform.
- Max Run - The max time taken to perform an inference for a given run.
- Min Run - The min time taken to perform an inference for a given run.
- Std - The Standard deviation between runs.
- Mean - The mean run time.
- Median - The median run time.
- IQR - The inter quartile range of the runs.
- Cores Logical - The number of logical cores on the host machine.
- Cores Physical - The number of physical Cores on the host machine.
Planned Future Stats
- Model Size - Information relating to the size of the model in bytes.
- GPU Stat Support - Information about if GPU is available and if it is being utilized.
Metadata
Release files for pyinfer 0.0.3
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| pyinfer-0.0.3.tar.gz | 9.3 kB | Without provenance | Details for pyinfer-0.0.3.tar.gz |
Release files / pyinfer-0.0.3.tar.gz
| Download URL | pyinfer-0.0.3.tar.gz |
|---|---|
| Size | 9.3 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
97aa21f9c0b65eb0527f68efcf88bde1c1b1a39958335a6eb82361cf85ccd273
|
|
BLAKE2b-256 checksum How to use checksums |
a59bf23ce1d8c12d4d88d8cbdc27823dfad6911e6f33a3414a3f466790605208
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/3.2.0 pkginfo/1.5.0.1 requests/2.24.0 setuptools/47.3.1.post20200622 requests-toolbelt/0.9.1 tqdm/4.47.0 CPython/3.7.7
|
Security for pyinfer 0.0.3
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.