AI Lab Dashboard
Model management, training, and inference monitoring
Active Models
7
3 in production
Training Jobs
2
1 queued
Inference/min
1,247
12ms avg latency
Datasets
14
2.4 TB total
Top Models by Usage
ewe-chat-7b847 req/min
sentiment-classifier234 req/min
product-recommender112 req/min
image-gen-xl54 req/min
Recent Activity
Training: fraud-detector-v3Epoch 47/100
Deployed: chat-7b to prod2h ago
Dataset: reviews-2026 uploaded5h ago
Alert: recommender latency spikeResolved
GPU Utilization
P40 #0
71% • 62°C • ewe-chat
P40 #1
89% • 71°C • training
P40 #2
42% • 55°C • image-gen
P40 #3
23% • 48°C • standby
Models
Manage deployed and in-development models
ewe-chat-7b Production
Large Language Model • Qwen 7B fine-tune
94.2%
Accuracy
12ms
Latency
847
Req/min
sentiment-classifier Production
Text Classification • DistilBERT fine-tune
97.1%
F1 Score
3ms
Latency
234
Req/min
fraud-detector-v3 Training
Anomaly Detection • Custom Transformer
Epoch 47/100 • Loss: 0.0234 • ETA: 3h 22m
image-gen-xl Production
Image Generation • SDXL fine-tune
0.87
FID Score
4.2s
Gen Time
54
Req/min
product-recommender Production
Recommendation • Two-Tower Neural
89.4%
Recall@10
8ms
Latency
112
Req/min
voice-to-text Draft
Speech Recognition • Whisper fine-tune
96.8%
WER
RTx0.3
Speed
-
Not deployed
Training Monitor
Live training job output
fraud-detector-v3 — Training
Epoch: 47/100
Loss: 0.0234
LR: 3e-5
GPU: P40 #1 (89%)
[Epoch 44] train_loss: 0.0289 val_loss: 0.0301 accuracy: 0.9847
[Epoch 45] train_loss: 0.0271 val_loss: 0.0284 accuracy: 0.9856
[Epoch 46] train_loss: 0.0252 val_loss: 0.0268 accuracy: 0.9862 LR decay triggered
[Epoch 47] train_loss: 0.0234 val_loss: 0.0249 accuracy: 0.9871
AI Playground
Test models interactively
Model:
Click "Run" to see model output
Datasets
Training and evaluation data management
| Name | Type | Size | Records | Last Updated | Used By |
|---|---|---|---|---|---|
| customer-reviews-2026 | Text | 1.2 GB | 4.2M | Feb 5 | sentiment-classifier |
| transaction-logs | Tabular | 340 MB | 12M | Feb 7 | fraud-detector |
| product-catalog | Structured | 89 MB | 850K | Feb 1 | product-recommender |
| chat-conversations | Text | 450 MB | 2.1M | Jan 28 | ewe-chat-7b |
| generated-images | Image | 28 GB | 120K | Feb 3 | image-gen-xl |
| voice-recordings | Audio | 15 GB | 45K | Jan 15 | voice-to-text |
Inference API
Deploy models as REST endpoints
Active Endpoints
| Endpoint | Model | Latency | Req/min | Status |
|---|---|---|---|---|
| /api/chat | ewe-chat-7b | 12ms | 847 | Online |
| /api/sentiment | sentiment-classifier | 3ms | 234 | Online |
| /api/recommend | product-recommender | 8ms | 112 | Online |
| /api/generate-image | image-gen-xl | 4.2s | 54 | Online |
Example Request
// Sentiment analysis
POST /api/sentiment
{ "text": "This product is amazing!" }
// Response
{ "label": "positive", "score": 0.987, "latency_ms": 3 }