mirrors
/
yolov5
mirror of https://github.com/ultralytics/yolov5.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Fix Actions CI

pull/13561/head
Glenn Jocher 2025-04-15 22:46:40 +02:00
parent 79355eb1b2
commit 61888d8fba
1 changed files with 12 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
utils/loggers/comet/README.md
View File

@ -37,8 +37,8 @@ You can configure Comet in two ways:
Set your credentials directly in your environment.
```shell
export COMET_API_KEY=<Your Comet API Key>
export COMET_PROJECT_NAME=<Your Comet Project Name> # Defaults to 'yolov5' if not set
export COMET_API_KEY=YOUR_COMET_API_KEY
export COMET_PROJECT_NAME=YOUR_COMET_PROJECT_NAME # Defaults to 'yolov5' if not set
```
Find your API key in your [Comet Account Settings](https://www.comet.com/site/).
@ -48,8 +48,8 @@ You can configure Comet in two ways:
```ini
[comet]
api_key=<Your Comet API Key>
project_name=<Your Comet Project Name> # Defaults to 'yolov5' if not set
api_key=YOUR_COMET_API_KEY
project_name=YOUR_COMET_PROJECT_NAME # Defaults to 'yolov5' if not set
```
### Run the Training Script
@ -103,11 +103,11 @@ Customize Comet's logging behavior using command-line flags or environment varia
```shell
# Environment Variables for Comet Configuration
export COMET_MODE=online # 'online' or 'offline'. Default: online
export COMET_MODEL_NAME=<your_model_name> # Name for the saved model. Default: yolov5
export COMET_MODEL_NAME=YOUR_MODEL_NAME # Name for the saved model. Default: yolov5
export COMET_LOG_CONFUSION_MATRIX=false # Disable confusion matrix logging. Default: true
export COMET_MAX_IMAGE_UPLOADS=<number> # Max prediction images to log. Default: 100
export COMET_MAX_IMAGE_UPLOADS=NUMBER # Max prediction images to log. Default: 100
export COMET_LOG_PER_CLASS_METRICS=true # Log metrics per class. Default: false
export COMET_DEFAULT_CHECKPOINT_FILENAME=<checkpoint_file.pt> # Checkpoint for resuming. Default: 'last.pt'
export COMET_DEFAULT_CHECKPOINT_FILENAME=checkpoint_file.pt # Checkpoint for resuming. Default: 'last.pt'
export COMET_LOG_BATCH_LEVEL_METRICS=true # Log training metrics per batch. Default: false
export COMET_LOG_PREDICTIONS=true # Disable prediction logging if set to false. Default: true
```
@ -208,7 +208,7 @@ To use a dataset stored in Comet Artifacts, update the `path` in your dataset YA
```yaml
# contents of artifact.yaml
path: "comet://<workspace_name>/<artifact_name>:<artifact_version_or_alias>"
path: "comet://WORKSPACE_NAME/ARTIFACT_NAME:ARTIFACT_VERSION_OR_ALIAS"
train: images/train # Adjust subdirectory if needed
val: images/val # Adjust subdirectory if needed
@ -231,13 +231,13 @@ Artifacts track data lineage, showing which experiments used specific dataset ve
## 🔄 Resuming Training Runs
If a training run is interrupted (for example, due to connection issues), you can resume it using the `--resume` flag with the Comet Run Path (`comet://<your_workspace>/<your_project>/<experiment_id>`).
If a training run is interrupted (for example, due to connection issues), you can resume it using the `--resume` flag with the Comet Run Path (`comet://YOUR_WORKSPACE/YOUR_PROJECT/EXPERIMENT_ID`).
This restores the model state, hyperparameters, arguments, and downloads necessary Artifacts, continuing logging to the existing Comet Experiment. Learn more about [resuming runs in the Comet documentation](https://www.comet.com/docs/v2/guides/experiment-management/resume-experiment/).
```shell
python train.py \
--resume "comet://<your_workspace>/<your_project>/<experiment_id>"
--resume "comet://YOUR_WORKSPACE/YOUR_PROJECT/EXPERIMENT_ID"
```
## 🔍 Hyperparameter Optimization (HPO)
@ -270,10 +270,10 @@ Execute multiple sweep trials concurrently using the `comet optimizer` command:
```shell
comet optimizer -j \
utils/loggers/comet/hpo.py < num_workers > utils/loggers/comet/optimizer_config.json
utils/loggers/comet/hpo.py NUM_WORKERS utils/loggers/comet/optimizer_config.json
```
Replace `<num_workers>` with the desired number of parallel processes.
Replace `NUM_WORKERS` with the desired number of parallel processes.
### Visualizing HPO Results