@@ -31,7 +31,6 @@ python3 .ci/scripts/benchmark_tooling/get_benchmark_analysis_data.py \
3131- ` --endTime ` : End time in ISO format (e.g., "2025-06-17T18:00:00") (required)
3232- ` --env ` : Choose environment ("local" or "prod", default: "prod")
3333- ` --no-silent ` : Show processing logs (default: only show results & minimum logging)
34- - ` print-all-table-info ` : show all cleaned table infos, this helps user to pick the correct format of filters
3534
3635##### Output Options:
3736- ` --outputType ` : Choose output format (default: "print")
@@ -49,15 +48,24 @@ python3 .ci/scripts/benchmark_tooling/get_benchmark_analysis_data.py \
4948- ` --models ` : Filter by specific model names (e.g "mv3" "meta-llama-llama-3.2-1b-instruct-qlora-int4-eo8")
5049
5150#### Example Usage
52- call multiple private device pools and models
51+ call multiple private device pools and models:
52+ this fetches all the private table data that has model ` llama-3.2-1B ` and ` mv3 `
5353``` bash
5454python3 get_benchmark_analysis_data.py \
5555--startTime " 2025-06-01T00:00:00" \
5656--endTime " 2025-06-11T00:00:00" \
57- --private-device-pools ' apple_iphone_15_private' ' samsung_s22_private'
58- --models ' mv3 ' ' allenai/OLMo- 1B-hf '
57+ --private-device-pools ' apple_iphone_15_private' ' samsung_s22_private' \
58+ --models ' meta-llama/Llama-3.2- 1B-Instruct-SpinQuant_INT4_EO8 ' ' mv3 '
5959```
6060
61+ this fetches all the private iphone table data that has model ` llama-3.2-1B ` and ` mv3 ` , and associated public iphone data
62+ ``` bash
63+ python3 get_benchmark_analysis_data.py \
64+ --startTime " 2025-06-01T00:00:00" \
65+ --endTime " 2025-06-11T00:00:00" \
66+ --private-device-pools ' apple_iphone_15_private' \
67+ --models ' meta-llama/Llama-3.2-1B-Instruct-SpinQuant_INT4_EO8' ' mv3'
68+ ```
6169#### Working with Output Files CSV and Excel
6270
6371You can use methods in ` common.py ` to convert the file data back to DataFrame format, those methods read the first row in csv/excel file, and return result with format list of {"groupInfo": DICT , "df": df .Dataframe{}} format.
0 commit comments