Skip to content

Commit 0da9878

Browse files
committed
separated tuner-specific args into their own section + message typo
Signed-off-by: Jeff Ng <[email protected]>
1 parent 5e33163 commit 0da9878

File tree

2 files changed

+39
-30
lines changed

2 files changed

+39
-30
lines changed

docs/user/InstructionsForAutoTuner.md

Lines changed: 38 additions & 29 deletions
Original file line numberDiff line numberDiff line change
@@ -154,35 +154,44 @@ GCP Setup Tutorial coming soon.
154154

155155

156156
### List of input arguments
157-
| Argument | Description |
158-
|-------------------------------|-------------------------------------------------------------------------------------------------------|
159-
| `--design` | Name of the design for Autotuning. |
160-
| `--platform` | Name of the platform for Autotuning. |
161-
| `--config` | Configuration file that sets which knobs to use for Autotuning. |
162-
| `--experiment` | Experiment name. This parameter is used to prefix the FLOW_VARIANT and to set the Ray log destination.|
163-
| `--resume` | Resume previous run. |
164-
| `--git_clean` | Clean binaries and build files. **WARNING**: may lose previous data. |
165-
| `--git_clone` | Force new git clone. **WARNING**: may lose previous data. |
166-
| `--git_clone_args` | Additional git clone arguments. |
167-
| `--git_latest` | Use latest version of OpenROAD app. |
168-
| `--git_or_branch` | OpenROAD app branch to use. |
169-
| `--git_orfs_branch` | OpenROAD-flow-scripts branch to use. |
170-
| `--git_url` | OpenROAD-flow-scripts repo URL to use. |
171-
| `--build_args` | Additional arguments given to ./build_openroad.sh |
172-
| `--algorithm` | Search algorithm to use for Autotuning. |
173-
| `--eval` | Evaluate function to use with search algorithm. |
174-
| `--samples` | Number of samples for tuning. |
175-
| `--iterations` | Number of iterations for tuning. |
176-
| `--resources_per_trial` | Number of CPUs to request for each tuning job. |
177-
| `--reference` | Reference file for use with PPAImprov. |
178-
| `--perturbation` | Perturbation interval for PopulationBasedTraining |
179-
| `--seed` | Random seed. |
180-
| `--jobs` | Max number of concurrent jobs. |
181-
| `--openroad_threads` | Max number of threads usable. |
182-
| `--server` | The address of Ray server to connect. |
183-
| `--port` | The port of Ray server to connect. |
184-
| `-v` or `--verbose` | Verbosity Level. [0: Only ray status, 1: print stderr, 2: print stdout on top of what is in level 0 and 1. ] |
185-
| | |
157+
| Argument | Description | Default |
158+
|-------------------------------|-------------------------------------------------------------------------------------------------------|---------|
159+
| `--design` | Name of the design for Autotuning. ||
160+
| `--platform` | Name of the platform for Autotuning. ||
161+
| `--config` | Configuration file that sets which knobs to use for Autotuning. ||
162+
| `--experiment` | Experiment name. This parameter is used to prefix the FLOW_VARIANT and to set the Ray log destination.| test |
163+
| `--git_clean` | Clean binaries and build files. **WARNING**: may lose previous data. ||
164+
| `--git_clone` | Force new git clone. **WARNING**: may lose previous data. ||
165+
| `--git_clone_args` | Additional git clone arguments. ||
166+
| `--git_latest` | Use latest version of OpenROAD app. ||
167+
| `--git_or_branch` | OpenROAD app branch to use. ||
168+
| `--git_orfs_branch` | OpenROAD-flow-scripts branch to use. ||
169+
| `--git_url` | OpenROAD-flow-scripts repo URL to use. | [ORFS GitHub repo](https://github.com/The-OpenROAD-Project/OpenROAD-flow-scripts) |
170+
| `--build_args` | Additional arguments given to ./build_openroad.sh ||
171+
| `--samples` | Number of samples for tuning. | 10 |
172+
| `--jobs` | Max number of concurrent jobs. | # of CPUs / 2 |
173+
| `--openroad_threads` | Max number of threads usable. | 16 |
174+
| `--server` | The address of Ray server to connect. ||
175+
| `--port` | The port of Ray server to connect. | 10001 |
176+
| `--timeout` | Time limit (in hours) for each trial run. | No limit |
177+
| `-v` or `--verbose` | Verbosity Level. [0: Only ray status, 1: print stderr, 2: print stdout on top of what is in level 0 and 1. ] | 0 |
178+
| | ||
179+
180+
#### Input arguments specific to tune mode
181+
The following input arguments are applicable for tune mode only.
182+
183+
| Argument | Description | Default |
184+
|-------------------------------|-------------------------------------------------------------------------------------------------------|---------|
185+
| `--algorithm` | Search algorithm to use for Autotuning. | hyperopt |
186+
| `--eval` | Evaluate function to use with search algorithm. ||
187+
| `--iterations` | Number of iterations for tuning. | 1 |
188+
| `--resources_per_trial` | Number of CPUs to request for each tuning job. | 1 |
189+
| `--reference` | Reference file for use with PPAImprov. ||
190+
| `--perturbation` | Perturbation interval for PopulationBasedTraining | 25 |
191+
| `--seed` | Random seed. | 42 |
192+
| `--resume` | Resume previous run. ||
193+
| | ||
194+
186195
### GUI
187196

188197
Basically, progress is displayed at the terminal where you run, and when all runs are finished, the results are displayed.

tools/AutoTuner/src/autotuner/distributed.py

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -789,7 +789,7 @@ def parse_arguments():
789789
type=int,
790790
metavar="<int>",
791791
default=1,
792-
help="Number of CPUs to request for each tunning job.",
792+
help="Number of CPUs to request for each tuning job.",
793793
)
794794
tune_parser.add_argument(
795795
"--reference",

0 commit comments

Comments
 (0)