opencode-skills-antigravity 1.0.39 → 1.0.41
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/bundled-skills/.antigravity-install-manifest.json +10 -1
- package/bundled-skills/docs/integrations/jetski-cortex.md +3 -3
- package/bundled-skills/docs/integrations/jetski-gemini-loader/README.md +1 -1
- package/bundled-skills/docs/maintainers/repo-growth-seo.md +3 -3
- package/bundled-skills/docs/maintainers/security-findings-triage-2026-03-29-refresh.csv +34 -0
- package/bundled-skills/docs/maintainers/security-findings-triage-2026-03-29-refresh.md +2 -0
- package/bundled-skills/docs/maintainers/skills-update-guide.md +1 -1
- package/bundled-skills/docs/sources/sources.md +2 -2
- package/bundled-skills/docs/users/bundles.md +1 -1
- package/bundled-skills/docs/users/claude-code-skills.md +1 -1
- package/bundled-skills/docs/users/gemini-cli-skills.md +1 -1
- package/bundled-skills/docs/users/getting-started.md +1 -1
- package/bundled-skills/docs/users/kiro-integration.md +1 -1
- package/bundled-skills/docs/users/usage.md +4 -4
- package/bundled-skills/docs/users/visual-guide.md +4 -4
- package/bundled-skills/hugging-face-cli/SKILL.md +192 -195
- package/bundled-skills/hugging-face-community-evals/SKILL.md +213 -0
- package/bundled-skills/hugging-face-community-evals/examples/.env.example +3 -0
- package/bundled-skills/hugging-face-community-evals/examples/USAGE_EXAMPLES.md +101 -0
- package/bundled-skills/hugging-face-community-evals/scripts/inspect_eval_uv.py +104 -0
- package/bundled-skills/hugging-face-community-evals/scripts/inspect_vllm_uv.py +306 -0
- package/bundled-skills/hugging-face-community-evals/scripts/lighteval_vllm_uv.py +297 -0
- package/bundled-skills/hugging-face-dataset-viewer/SKILL.md +120 -120
- package/bundled-skills/hugging-face-gradio/SKILL.md +304 -0
- package/bundled-skills/hugging-face-gradio/examples.md +613 -0
- package/bundled-skills/hugging-face-jobs/SKILL.md +25 -18
- package/bundled-skills/hugging-face-jobs/index.html +216 -0
- package/bundled-skills/hugging-face-jobs/references/hardware_guide.md +336 -0
- package/bundled-skills/hugging-face-jobs/references/hub_saving.md +352 -0
- package/bundled-skills/hugging-face-jobs/references/token_usage.md +570 -0
- package/bundled-skills/hugging-face-jobs/references/troubleshooting.md +475 -0
- package/bundled-skills/hugging-face-jobs/scripts/cot-self-instruct.py +718 -0
- package/bundled-skills/hugging-face-jobs/scripts/finepdfs-stats.py +546 -0
- package/bundled-skills/hugging-face-jobs/scripts/generate-responses.py +587 -0
- package/bundled-skills/hugging-face-model-trainer/SKILL.md +11 -12
- package/bundled-skills/hugging-face-model-trainer/references/gguf_conversion.md +296 -0
- package/bundled-skills/hugging-face-model-trainer/references/hardware_guide.md +283 -0
- package/bundled-skills/hugging-face-model-trainer/references/hub_saving.md +364 -0
- package/bundled-skills/hugging-face-model-trainer/references/local_training_macos.md +231 -0
- package/bundled-skills/hugging-face-model-trainer/references/reliability_principles.md +371 -0
- package/bundled-skills/hugging-face-model-trainer/references/trackio_guide.md +189 -0
- package/bundled-skills/hugging-face-model-trainer/references/training_methods.md +150 -0
- package/bundled-skills/hugging-face-model-trainer/references/training_patterns.md +203 -0
- package/bundled-skills/hugging-face-model-trainer/references/troubleshooting.md +282 -0
- package/bundled-skills/hugging-face-model-trainer/references/unsloth.md +313 -0
- package/bundled-skills/hugging-face-model-trainer/scripts/convert_to_gguf.py +424 -0
- package/bundled-skills/hugging-face-model-trainer/scripts/dataset_inspector.py +417 -0
- package/bundled-skills/hugging-face-model-trainer/scripts/estimate_cost.py +150 -0
- package/bundled-skills/hugging-face-model-trainer/scripts/train_dpo_example.py +106 -0
- package/bundled-skills/hugging-face-model-trainer/scripts/train_grpo_example.py +89 -0
- package/bundled-skills/hugging-face-model-trainer/scripts/train_sft_example.py +122 -0
- package/bundled-skills/hugging-face-model-trainer/scripts/unsloth_sft_example.py +512 -0
- package/bundled-skills/hugging-face-paper-publisher/SKILL.md +11 -4
- package/bundled-skills/hugging-face-paper-publisher/examples/example_usage.md +326 -0
- package/bundled-skills/hugging-face-paper-publisher/references/quick_reference.md +216 -0
- package/bundled-skills/hugging-face-paper-publisher/scripts/paper_manager.py +606 -0
- package/bundled-skills/hugging-face-paper-publisher/templates/arxiv.md +299 -0
- package/bundled-skills/hugging-face-paper-publisher/templates/ml-report.md +358 -0
- package/bundled-skills/hugging-face-paper-publisher/templates/modern.md +319 -0
- package/bundled-skills/hugging-face-paper-publisher/templates/standard.md +201 -0
- package/bundled-skills/hugging-face-papers/SKILL.md +241 -0
- package/bundled-skills/hugging-face-trackio/.claude-plugin/plugin.json +19 -0
- package/bundled-skills/hugging-face-trackio/SKILL.md +117 -0
- package/bundled-skills/hugging-face-trackio/references/alerts.md +196 -0
- package/bundled-skills/hugging-face-trackio/references/logging_metrics.md +206 -0
- package/bundled-skills/hugging-face-trackio/references/retrieving_metrics.md +251 -0
- package/bundled-skills/hugging-face-vision-trainer/SKILL.md +595 -0
- package/bundled-skills/hugging-face-vision-trainer/references/finetune_sam2_trainer.md +254 -0
- package/bundled-skills/hugging-face-vision-trainer/references/hub_saving.md +618 -0
- package/bundled-skills/hugging-face-vision-trainer/references/image_classification_training_notebook.md +279 -0
- package/bundled-skills/hugging-face-vision-trainer/references/object_detection_training_notebook.md +700 -0
- package/bundled-skills/hugging-face-vision-trainer/references/reliability_principles.md +310 -0
- package/bundled-skills/hugging-face-vision-trainer/references/timm_trainer.md +91 -0
- package/bundled-skills/hugging-face-vision-trainer/scripts/dataset_inspector.py +814 -0
- package/bundled-skills/hugging-face-vision-trainer/scripts/estimate_cost.py +217 -0
- package/bundled-skills/hugging-face-vision-trainer/scripts/image_classification_training.py +383 -0
- package/bundled-skills/hugging-face-vision-trainer/scripts/object_detection_training.py +710 -0
- package/bundled-skills/hugging-face-vision-trainer/scripts/sam_segmentation_training.py +382 -0
- package/bundled-skills/jq/SKILL.md +273 -0
- package/bundled-skills/odoo-edi-connector/SKILL.md +32 -10
- package/bundled-skills/odoo-woocommerce-bridge/SKILL.md +9 -5
- package/bundled-skills/tmux/SKILL.md +370 -0
- package/bundled-skills/transformers-js/SKILL.md +639 -0
- package/bundled-skills/transformers-js/references/CACHE.md +339 -0
- package/bundled-skills/transformers-js/references/CONFIGURATION.md +390 -0
- package/bundled-skills/transformers-js/references/EXAMPLES.md +605 -0
- package/bundled-skills/transformers-js/references/MODEL_ARCHITECTURES.md +167 -0
- package/bundled-skills/transformers-js/references/PIPELINE_OPTIONS.md +545 -0
- package/bundled-skills/transformers-js/references/TEXT_GENERATION.md +315 -0
- package/bundled-skills/viboscope/SKILL.md +64 -0
- package/package.json +1 -1
|
@@ -0,0 +1,206 @@
|
|
|
1
|
+
# Logging Metrics with Trackio
|
|
2
|
+
|
|
3
|
+
**Trackio** is a lightweight, free experiment tracking library from Hugging Face. It provides a wandb-compatible API for logging metrics with local-first design.
|
|
4
|
+
|
|
5
|
+
- **GitHub**: [gradio-app/trackio](https://github.com/gradio-app/trackio)
|
|
6
|
+
- **Docs**: [huggingface.co/docs/trackio](https://huggingface.co/docs/trackio/index)
|
|
7
|
+
|
|
8
|
+
## Installation
|
|
9
|
+
|
|
10
|
+
```bash
|
|
11
|
+
pip install trackio
|
|
12
|
+
# or
|
|
13
|
+
uv pip install trackio
|
|
14
|
+
```
|
|
15
|
+
|
|
16
|
+
## Core API
|
|
17
|
+
|
|
18
|
+
### Basic Usage
|
|
19
|
+
|
|
20
|
+
```python
|
|
21
|
+
import trackio
|
|
22
|
+
|
|
23
|
+
# Initialize a run
|
|
24
|
+
trackio.init(
|
|
25
|
+
project="my-project",
|
|
26
|
+
config={"learning_rate": 0.001, "epochs": 10}
|
|
27
|
+
)
|
|
28
|
+
|
|
29
|
+
# Log metrics during training
|
|
30
|
+
for epoch in range(10):
|
|
31
|
+
loss = train_epoch()
|
|
32
|
+
trackio.log({"loss": loss, "epoch": epoch})
|
|
33
|
+
|
|
34
|
+
# Finalize the run
|
|
35
|
+
trackio.finish()
|
|
36
|
+
```
|
|
37
|
+
|
|
38
|
+
### Key Functions
|
|
39
|
+
|
|
40
|
+
| Function | Purpose |
|
|
41
|
+
|----------|---------|
|
|
42
|
+
| `trackio.init(...)` | Start a new tracking run |
|
|
43
|
+
| `trackio.log(dict)` | Log metrics (called repeatedly during training) |
|
|
44
|
+
| `trackio.finish()` | Finalize run and ensure all metrics are saved |
|
|
45
|
+
| `trackio.show()` | Launch the local dashboard |
|
|
46
|
+
| `trackio.sync(...)` | Sync local project to HF Space |
|
|
47
|
+
|
|
48
|
+
## trackio.init() Parameters
|
|
49
|
+
|
|
50
|
+
```python
|
|
51
|
+
trackio.init(
|
|
52
|
+
project="my-project", # Project name (groups runs together)
|
|
53
|
+
name="run-name", # Optional: name for this specific run
|
|
54
|
+
config={...}, # Hyperparameters and config to log
|
|
55
|
+
space_id="username/trackio", # Optional: sync to HF Space for remote dashboard
|
|
56
|
+
group="experiment-group", # Optional: group related runs
|
|
57
|
+
)
|
|
58
|
+
```
|
|
59
|
+
|
|
60
|
+
## Local vs Remote Dashboard
|
|
61
|
+
|
|
62
|
+
### Local (Default)
|
|
63
|
+
|
|
64
|
+
By default, trackio stores metrics in a local SQLite database and runs the dashboard locally:
|
|
65
|
+
|
|
66
|
+
```python
|
|
67
|
+
trackio.init(project="my-project")
|
|
68
|
+
# ... training ...
|
|
69
|
+
trackio.finish()
|
|
70
|
+
|
|
71
|
+
# Launch local dashboard
|
|
72
|
+
trackio.show()
|
|
73
|
+
```
|
|
74
|
+
|
|
75
|
+
Or from terminal:
|
|
76
|
+
```bash
|
|
77
|
+
trackio show --project my-project
|
|
78
|
+
```
|
|
79
|
+
|
|
80
|
+
### Remote (HF Space)
|
|
81
|
+
|
|
82
|
+
Pass `space_id` to sync metrics to a Hugging Face Space for persistent, shareable dashboards:
|
|
83
|
+
|
|
84
|
+
```python
|
|
85
|
+
trackio.init(
|
|
86
|
+
project="my-project",
|
|
87
|
+
space_id="username/trackio" # Auto-creates Space if it doesn't exist
|
|
88
|
+
)
|
|
89
|
+
```
|
|
90
|
+
|
|
91
|
+
⚠️ **For remote training** (cloud GPUs, HF Jobs, etc.): Always use `space_id` since local storage is lost when the instance terminates.
|
|
92
|
+
|
|
93
|
+
### Sync Local to Remote
|
|
94
|
+
|
|
95
|
+
Sync existing local projects to a Space:
|
|
96
|
+
|
|
97
|
+
```python
|
|
98
|
+
trackio.sync(project="my-project", space_id="username/my-experiments")
|
|
99
|
+
```
|
|
100
|
+
|
|
101
|
+
## wandb Compatibility
|
|
102
|
+
|
|
103
|
+
Trackio is API-compatible with wandb. Drop-in replacement:
|
|
104
|
+
|
|
105
|
+
```python
|
|
106
|
+
import trackio as wandb
|
|
107
|
+
|
|
108
|
+
wandb.init(project="my-project")
|
|
109
|
+
wandb.log({"loss": 0.5})
|
|
110
|
+
wandb.finish()
|
|
111
|
+
```
|
|
112
|
+
|
|
113
|
+
## TRL Integration
|
|
114
|
+
|
|
115
|
+
When using TRL trainers, set `report_to="trackio"` for automatic metric logging:
|
|
116
|
+
|
|
117
|
+
```python
|
|
118
|
+
from trl import SFTConfig, SFTTrainer
|
|
119
|
+
import trackio
|
|
120
|
+
|
|
121
|
+
trackio.init(
|
|
122
|
+
project="sft-training",
|
|
123
|
+
space_id="username/trackio",
|
|
124
|
+
config={"model": "Qwen/Qwen2.5-0.5B", "dataset": "trl-lib/Capybara"}
|
|
125
|
+
)
|
|
126
|
+
|
|
127
|
+
config = SFTConfig(
|
|
128
|
+
output_dir="./output",
|
|
129
|
+
report_to="trackio", # Automatic metric logging
|
|
130
|
+
# ... other config
|
|
131
|
+
)
|
|
132
|
+
|
|
133
|
+
trainer = SFTTrainer(model=model, args=config, ...)
|
|
134
|
+
trainer.train()
|
|
135
|
+
trackio.finish()
|
|
136
|
+
```
|
|
137
|
+
|
|
138
|
+
## What Gets Logged
|
|
139
|
+
|
|
140
|
+
With TRL/Transformers integration, trackio automatically captures:
|
|
141
|
+
- Training loss
|
|
142
|
+
- Learning rate
|
|
143
|
+
- Eval metrics
|
|
144
|
+
- Training throughput
|
|
145
|
+
|
|
146
|
+
For manual logging, log any numeric metrics:
|
|
147
|
+
|
|
148
|
+
```python
|
|
149
|
+
trackio.log({
|
|
150
|
+
"train_loss": 0.5,
|
|
151
|
+
"train_accuracy": 0.85,
|
|
152
|
+
"val_loss": 0.4,
|
|
153
|
+
"val_accuracy": 0.88,
|
|
154
|
+
"epoch": 1
|
|
155
|
+
})
|
|
156
|
+
```
|
|
157
|
+
|
|
158
|
+
## Grouping Runs
|
|
159
|
+
|
|
160
|
+
Use `group` to organize related experiments in the dashboard sidebar:
|
|
161
|
+
|
|
162
|
+
```python
|
|
163
|
+
# Group by experiment type
|
|
164
|
+
trackio.init(project="my-project", name="baseline-v1", group="baseline")
|
|
165
|
+
trackio.init(project="my-project", name="augmented-v1", group="augmented")
|
|
166
|
+
|
|
167
|
+
# Group by hyperparameter
|
|
168
|
+
trackio.init(project="hyperparam-sweep", name="lr-0.001", group="lr_0.001")
|
|
169
|
+
trackio.init(project="hyperparam-sweep", name="lr-0.01", group="lr_0.01")
|
|
170
|
+
```
|
|
171
|
+
|
|
172
|
+
## Configuration Best Practices
|
|
173
|
+
|
|
174
|
+
Keep config minimal — only log what's useful for comparing runs:
|
|
175
|
+
|
|
176
|
+
```python
|
|
177
|
+
trackio.init(
|
|
178
|
+
project="qwen-sft-capybara",
|
|
179
|
+
name="baseline-lr2e5",
|
|
180
|
+
config={
|
|
181
|
+
"model": "Qwen/Qwen2.5-0.5B",
|
|
182
|
+
"dataset": "trl-lib/Capybara",
|
|
183
|
+
"learning_rate": 2e-5,
|
|
184
|
+
"num_epochs": 3,
|
|
185
|
+
"batch_size": 8,
|
|
186
|
+
}
|
|
187
|
+
)
|
|
188
|
+
```
|
|
189
|
+
|
|
190
|
+
## Embedding Dashboards
|
|
191
|
+
|
|
192
|
+
Embed Space dashboards in websites with query parameters:
|
|
193
|
+
|
|
194
|
+
```html
|
|
195
|
+
<iframe
|
|
196
|
+
src="https://username-trackio.hf.space/?project=my-project&metrics=train_loss,val_loss&sidebar=hidden"
|
|
197
|
+
style="width:1600px; height:500px; border:0;">
|
|
198
|
+
</iframe>
|
|
199
|
+
```
|
|
200
|
+
|
|
201
|
+
Query parameters:
|
|
202
|
+
- `project`: Filter to specific project
|
|
203
|
+
- `metrics`: Comma-separated metric names to show
|
|
204
|
+
- `sidebar`: `hidden` or `collapsed`
|
|
205
|
+
- `smoothing`: 0-20 (smoothing slider value)
|
|
206
|
+
- `xmin`, `xmax`: X-axis limits
|
|
@@ -0,0 +1,251 @@
|
|
|
1
|
+
# Retrieving Metrics with Trackio CLI
|
|
2
|
+
|
|
3
|
+
The `trackio` CLI provides direct terminal access to query Trackio experiment tracking data locally without needing to start the MCP server.
|
|
4
|
+
|
|
5
|
+
## Quick Command Reference
|
|
6
|
+
|
|
7
|
+
| Task | Command |
|
|
8
|
+
|------|---------|
|
|
9
|
+
| List projects | `trackio list projects` |
|
|
10
|
+
| List runs | `trackio list runs --project <name>` |
|
|
11
|
+
| List metrics | `trackio list metrics --project <name> --run <name>` |
|
|
12
|
+
| List system metrics | `trackio list system-metrics --project <name> --run <name>` |
|
|
13
|
+
| List alerts | `trackio list alerts --project <name> [--run <name>] [--level <level>] [--since <timestamp>]` |
|
|
14
|
+
| Get project summary | `trackio get project --project <name>` |
|
|
15
|
+
| Get run summary | `trackio get run --project <name> --run <name>` |
|
|
16
|
+
| Get metric values | `trackio get metric --project <name> --run <name> --metric <name>` |
|
|
17
|
+
| Get metric at step | `trackio get metric ... --metric <name> --step <N>` |
|
|
18
|
+
| Get metric around step | `trackio get metric ... --metric <name> --around <N> --window <W>` |
|
|
19
|
+
| Get all metrics snapshot | `trackio get snapshot --project <name> --run <name> --step <N>` |
|
|
20
|
+
| Get system metrics | `trackio get system-metric --project <name> --run <name>` |
|
|
21
|
+
| Show dashboard | `trackio show [--project <name>]` |
|
|
22
|
+
| Sync to Space | `trackio sync --project <name> --space-id <space_id>` |
|
|
23
|
+
|
|
24
|
+
## Core Commands
|
|
25
|
+
|
|
26
|
+
### List Commands
|
|
27
|
+
|
|
28
|
+
```bash
|
|
29
|
+
trackio list projects # List all projects
|
|
30
|
+
trackio list projects --json # JSON output
|
|
31
|
+
|
|
32
|
+
trackio list runs --project <name> # List runs in project
|
|
33
|
+
trackio list runs --project <name> --json # JSON output
|
|
34
|
+
|
|
35
|
+
trackio list metrics --project <name> --run <name> # List metrics for run
|
|
36
|
+
trackio list metrics --project <name> --run <name> --json
|
|
37
|
+
|
|
38
|
+
trackio list system-metrics --project <name> --run <name> # List system metrics
|
|
39
|
+
trackio list system-metrics --project <name> --run <name> --json
|
|
40
|
+
|
|
41
|
+
trackio list alerts --project <name> # List alerts
|
|
42
|
+
trackio list alerts --project <name> --run <name> --json # Filter by run
|
|
43
|
+
trackio list alerts --project <name> --level error --json # Filter by level
|
|
44
|
+
trackio list alerts --project <name> --json --since <ts> # Poll since timestamp
|
|
45
|
+
```
|
|
46
|
+
|
|
47
|
+
### Get Commands
|
|
48
|
+
|
|
49
|
+
```bash
|
|
50
|
+
trackio get project --project <name> # Project summary
|
|
51
|
+
trackio get project --project <name> --json # JSON output
|
|
52
|
+
|
|
53
|
+
trackio get run --project <name> --run <name> # Run summary
|
|
54
|
+
trackio get run --project <name> --run <name> --json
|
|
55
|
+
|
|
56
|
+
trackio get metric --project <name> --run <name> --metric <name> # Metric values
|
|
57
|
+
trackio get metric --project <name> --run <name> --metric <name> --json
|
|
58
|
+
trackio get metric ... --metric <name> --step 200 # At exact step
|
|
59
|
+
trackio get metric ... --metric <name> --around 200 --window 10 # ±10 steps
|
|
60
|
+
trackio get metric ... --metric <name> --at-time <ts> --window 60 # ±60 seconds
|
|
61
|
+
|
|
62
|
+
trackio get snapshot --project <name> --run <name> --step 200 --json # All metrics at step
|
|
63
|
+
trackio get snapshot --project <name> --run <name> --around 200 --window 5 --json # Window
|
|
64
|
+
trackio get snapshot --project <name> --run <name> --at-time <ts> --window 60 --json
|
|
65
|
+
|
|
66
|
+
trackio get system-metric --project <name> --run <name> # All system metrics
|
|
67
|
+
trackio get system-metric --project <name> --run <name> --metric <name> # Specific metric
|
|
68
|
+
trackio get system-metric --project <name> --run <name> --json
|
|
69
|
+
```
|
|
70
|
+
|
|
71
|
+
### Dashboard Commands
|
|
72
|
+
|
|
73
|
+
```bash
|
|
74
|
+
trackio show # Launch dashboard
|
|
75
|
+
trackio show --project <name> # Load specific project
|
|
76
|
+
trackio show --theme <theme> # Custom theme
|
|
77
|
+
trackio show --mcp-server # Enable MCP server
|
|
78
|
+
trackio show --color-palette "#FF0000,#00FF00" # Custom colors
|
|
79
|
+
```
|
|
80
|
+
|
|
81
|
+
### Sync Commands
|
|
82
|
+
|
|
83
|
+
```bash
|
|
84
|
+
trackio sync --project <name> --space-id <space_id> # Sync to HF Space
|
|
85
|
+
trackio sync --project <name> --space-id <space_id> --private # Private space
|
|
86
|
+
trackio sync --project <name> --space-id <space_id> --force # Overwrite
|
|
87
|
+
```
|
|
88
|
+
|
|
89
|
+
## Output Formats
|
|
90
|
+
|
|
91
|
+
All `list` and `get` commands support two output formats:
|
|
92
|
+
|
|
93
|
+
- **Human-readable** (default): Formatted text for terminal viewing
|
|
94
|
+
- **JSON** (with `--json` flag): Structured JSON for programmatic use
|
|
95
|
+
|
|
96
|
+
## Common Patterns
|
|
97
|
+
|
|
98
|
+
### Discover Projects and Runs
|
|
99
|
+
|
|
100
|
+
```bash
|
|
101
|
+
# List all available projects
|
|
102
|
+
trackio list projects
|
|
103
|
+
|
|
104
|
+
# List runs in a project
|
|
105
|
+
trackio list runs --project my-project
|
|
106
|
+
|
|
107
|
+
# Get project overview
|
|
108
|
+
trackio get project --project my-project --json
|
|
109
|
+
```
|
|
110
|
+
|
|
111
|
+
### Inspect Run Details
|
|
112
|
+
|
|
113
|
+
```bash
|
|
114
|
+
# Get run summary with all metrics
|
|
115
|
+
trackio get run --project my-project --run my-run --json
|
|
116
|
+
|
|
117
|
+
# List available metrics
|
|
118
|
+
trackio list metrics --project my-project --run my-run
|
|
119
|
+
|
|
120
|
+
# Get specific metric values
|
|
121
|
+
trackio get metric --project my-project --run my-run --metric loss --json
|
|
122
|
+
```
|
|
123
|
+
|
|
124
|
+
### Query System Metrics
|
|
125
|
+
|
|
126
|
+
```bash
|
|
127
|
+
# List system metrics (GPU, etc.)
|
|
128
|
+
trackio list system-metrics --project my-project --run my-run
|
|
129
|
+
|
|
130
|
+
# Get all system metric data
|
|
131
|
+
trackio get system-metric --project my-project --run my-run --json
|
|
132
|
+
|
|
133
|
+
# Get specific system metric
|
|
134
|
+
trackio get system-metric --project my-project --run my-run --metric gpu_utilization --json
|
|
135
|
+
```
|
|
136
|
+
|
|
137
|
+
### Automation Scripts
|
|
138
|
+
|
|
139
|
+
```bash
|
|
140
|
+
# Extract latest metric value
|
|
141
|
+
LATEST_LOSS=$(trackio get metric --project my-project --run my-run --metric loss --json | jq -r '.values[-1].value')
|
|
142
|
+
|
|
143
|
+
# Export run summary to file
|
|
144
|
+
trackio get run --project my-project --run my-run --json > run_summary.json
|
|
145
|
+
|
|
146
|
+
# Filter runs with jq
|
|
147
|
+
trackio list runs --project my-project --json | jq '.runs[] | select(startswith("train"))'
|
|
148
|
+
```
|
|
149
|
+
|
|
150
|
+
### LLM Agent Workflow
|
|
151
|
+
|
|
152
|
+
```bash
|
|
153
|
+
# 1. Discover available projects
|
|
154
|
+
trackio list projects --json
|
|
155
|
+
|
|
156
|
+
# 2. Explore project structure
|
|
157
|
+
trackio get project --project my-project --json
|
|
158
|
+
|
|
159
|
+
# 3. Inspect specific run
|
|
160
|
+
trackio get run --project my-project --run my-run --json
|
|
161
|
+
|
|
162
|
+
# 4. Query metric values
|
|
163
|
+
trackio get metric --project my-project --run my-run --metric accuracy --json
|
|
164
|
+
|
|
165
|
+
# 5. Poll for alerts (use --since for efficient incremental polling)
|
|
166
|
+
trackio list alerts --project my-project --json --since "2025-06-01T00:00:00"
|
|
167
|
+
|
|
168
|
+
# 6. When an alert fires at step N, get all metrics around that point
|
|
169
|
+
trackio get snapshot --project my-project --run my-run --around 200 --window 5 --json
|
|
170
|
+
```
|
|
171
|
+
|
|
172
|
+
## Error Handling
|
|
173
|
+
|
|
174
|
+
Commands validate inputs and return clear errors:
|
|
175
|
+
|
|
176
|
+
- Missing project: `Error: Project '<name>' not found.`
|
|
177
|
+
- Missing run: `Error: Run '<name>' not found in project '<project>'.`
|
|
178
|
+
- Missing metric: `Error: Metric '<name>' not found in run '<run>' of project '<project>'.`
|
|
179
|
+
|
|
180
|
+
All errors exit with non-zero status code and write to stderr.
|
|
181
|
+
|
|
182
|
+
## Key Options
|
|
183
|
+
|
|
184
|
+
- `--project`: Project name (required for most commands)
|
|
185
|
+
- `--run`: Run name (required for run-specific commands)
|
|
186
|
+
- `--metric`: Metric name (required for metric-specific commands)
|
|
187
|
+
- `--json`: Output in JSON format instead of human-readable
|
|
188
|
+
- `--step`: Exact step filter (for `get metric`, `get snapshot`)
|
|
189
|
+
- `--around`: Center step for window filter (for `get metric`, `get snapshot`)
|
|
190
|
+
- `--at-time`: Center ISO timestamp for window filter (for `get metric`, `get snapshot`)
|
|
191
|
+
- `--window`: Window size: ±steps for `--around`, ±seconds for `--at-time` (default: 10)
|
|
192
|
+
- `--level`: Alert level filter (`info`, `warn`, `error`) (for `list alerts`)
|
|
193
|
+
- `--since`: ISO timestamp to filter alerts after (for `list alerts`)
|
|
194
|
+
- `--theme`: Dashboard theme (for `show` command)
|
|
195
|
+
- `--mcp-server`: Enable MCP server mode (for `show` command)
|
|
196
|
+
- `--color-palette`: Comma-separated hex colors (for `show` command)
|
|
197
|
+
- `--private`: Create private Space (for `sync` command)
|
|
198
|
+
- `--force`: Overwrite existing database (for `sync` command)
|
|
199
|
+
|
|
200
|
+
## JSON Output Structure
|
|
201
|
+
|
|
202
|
+
### List Projects
|
|
203
|
+
```json
|
|
204
|
+
{"projects": ["project1", "project2"]}
|
|
205
|
+
```
|
|
206
|
+
|
|
207
|
+
### List Runs
|
|
208
|
+
```json
|
|
209
|
+
{"project": "my-project", "runs": ["run1", "run2"]}
|
|
210
|
+
```
|
|
211
|
+
|
|
212
|
+
### Project Summary
|
|
213
|
+
```json
|
|
214
|
+
{
|
|
215
|
+
"project": "my-project",
|
|
216
|
+
"num_runs": 3,
|
|
217
|
+
"runs": ["run1", "run2", "run3"],
|
|
218
|
+
"last_activity": 100
|
|
219
|
+
}
|
|
220
|
+
```
|
|
221
|
+
|
|
222
|
+
### Run Summary
|
|
223
|
+
```json
|
|
224
|
+
{
|
|
225
|
+
"project": "my-project",
|
|
226
|
+
"run": "my-run",
|
|
227
|
+
"num_logs": 50,
|
|
228
|
+
"metrics": ["loss", "accuracy"],
|
|
229
|
+
"config": {"learning_rate": 0.001},
|
|
230
|
+
"last_step": 49
|
|
231
|
+
}
|
|
232
|
+
```
|
|
233
|
+
|
|
234
|
+
### Metric Values
|
|
235
|
+
```json
|
|
236
|
+
{
|
|
237
|
+
"project": "my-project",
|
|
238
|
+
"run": "my-run",
|
|
239
|
+
"metric": "loss",
|
|
240
|
+
"values": [
|
|
241
|
+
{"step": 0, "timestamp": "2024-01-01T00:00:00", "value": 0.5},
|
|
242
|
+
{"step": 1, "timestamp": "2024-01-01T00:01:00", "value": 0.4}
|
|
243
|
+
]
|
|
244
|
+
}
|
|
245
|
+
```
|
|
246
|
+
|
|
247
|
+
## References
|
|
248
|
+
|
|
249
|
+
- **Complete CLI documentation**: See [docs/source/cli_commands.md](docs/source/cli_commands.md)
|
|
250
|
+
- **API and MCP Server**: See [docs/source/api_mcp_server.md](docs/source/api_mcp_server.md)
|
|
251
|
+
|