Update README.md
Browse files
README.md
CHANGED
|
@@ -149,11 +149,13 @@ This SFT approach enables Alpie-Core to deliver reliable, aligned, and context-a
|
|
| 149 |
|
| 150 |
**Carbon Footprint**: We estimated the environmental impact of training Alpie-Core (32B) on 8× NVIDIA A100-80GB GPUs (Azure) by calculating carbon emissions from GPU energy consumption. The calculation follows the formula:
|
| 151 |
CO₂e (kg) = Grid CO₂ Factor (kg/kWh) × Runtime (hours) × Power per GPU (kW) × Number of GPUs
|
|
|
|
| 152 |
Training Parameters:
|
| 153 |
Grid CO₂ Factor (Azure average): 0.364 kg CO₂e per kWh (source: Microsoft Sustainability Report)
|
| 154 |
Runtime: 408 hours
|
| 155 |
GPUs: 8× A100-80GB
|
| 156 |
We report results under two assumption modes:
|
|
|
|
| 157 |
Realistic mode (average training draw ≈ 250 W per GPU = 0.25 kWh/hr): 0.364 × 408 × 0.25 × 8 ≈ 298 kg CO₂e
|
| 158 |
|
| 159 |
|
|
|
|
| 149 |
|
| 150 |
**Carbon Footprint**: We estimated the environmental impact of training Alpie-Core (32B) on 8× NVIDIA A100-80GB GPUs (Azure) by calculating carbon emissions from GPU energy consumption. The calculation follows the formula:
|
| 151 |
CO₂e (kg) = Grid CO₂ Factor (kg/kWh) × Runtime (hours) × Power per GPU (kW) × Number of GPUs
|
| 152 |
+
|
| 153 |
Training Parameters:
|
| 154 |
Grid CO₂ Factor (Azure average): 0.364 kg CO₂e per kWh (source: Microsoft Sustainability Report)
|
| 155 |
Runtime: 408 hours
|
| 156 |
GPUs: 8× A100-80GB
|
| 157 |
We report results under two assumption modes:
|
| 158 |
+
|
| 159 |
Realistic mode (average training draw ≈ 250 W per GPU = 0.25 kWh/hr): 0.364 × 408 × 0.25 × 8 ≈ 298 kg CO₂e
|
| 160 |
|
| 161 |
|