TODO: Write a project description
1. Fork it!
2. Create your feature branch: `git checkout -b my-new-feature`
3. Commit your changes: `git commit -am 'Add some feature'`
4. Push to the branch: `git push origin my-new-feature`
5. Submit a pull request :D
-
1 Get the repository with git clone https://github.com/fernandoFernandeSantos/radiation-benchmarks-parsers.git
-
2 Set parameters for each particular benchmark available in https://github.com/dagoliveira/radiation-benchmarks.git Note: for some benchmarks such as Py-Faster-Rcnn and Darknet, you must have the Golden output for each platform, for error criticality evaluation
Before use ParseBenchmarkCriticality.py keep in mind that all benchmarks parameters must be set in Parameters.py, on the contrary, this parser will crash.
usage:
<path>/ParseBenchmarksCriticality.py [-h] --gen_database [GEN_DATA] --out_database [OUT_DATA] --database [ERROR_DATABASE] --benchmarks [BENCHMARKS] --parse_layers --pr_threshold [PR_THRESHOLD] --check_csv --ecc --is_fi
Parse logs for Neural Networks
arguments:
-h, --help show this help message and exit
--gen_database <GEN_DATA>
If this flag is passed, the other flags will have no
effects, despite out_database.
--gen_data <path where the parser must search
for ALL LOGs FILES
--out_database <OUT_DATA>
The output database name. If gen_database is used,
this flag will set the filename for the generated database.
The default filename is ./error_log_database.
--database <ERROR_DATABASE>
Where database is located
--benchmarks
A list separated by ',' (commas with no space) where
each item will be the benchmarks that parser will
process. Available parsers:
* Darknet --> needs --parse_layers and a Precision and Recall threshold value, only for DarknetV1 logs from 2016.
* DarknetV1 --> needs --parse_layers and a Precision and Recall threshold value. For DarknetV1 logs after 2017
* DarknetV2 --> needs --parse_layers and a Precision and Recall threshold value.
* Lenet --> needs --parse_layers and a Precision and Recall threshold value.
* Hotspot
* GEMM,
* HOG --> needs a Precision and Recall threshold value.
* lavamd
* nw
* quicksort
* accl
* PyFasterRCNN --> needs a Precision and Recall threshold value.
* Lulesh
* LUD
* mergesort
--parse_layers
If you want to parse Darknet layers, set it True, default
values is False
--pr_threshold <PR_THRESHOLD>
Precision and Recall threshold value,0 - 1, default
value is 0.5
--check_csv
This parameter will open a csv file which contains all
radiation test runs, then it will check if every SDC
is on a valid run, the default is false
--ecc
If the boards have ecc this is passed, otherwise
nothing must be passed
--is_fi
if it is a fault injection log processing
--err_hist
This parameter will generate an histogram for a serie of error thresholds,
these error thresholds are calculated using ERROR_RELATIVE_HISTOGRAM dict values
(set on Parameters.py)
TODO: Write history
TODO: Write credits
Copyright 2017 UFRGS HPC Reliability Group
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.