Add validation artifact validation/benchmark/qwen_image_bf16_vs_fp8_fixed_1024_8_profile_compare.md
Browse files
validation/benchmark/qwen_image_bf16_vs_fp8_fixed_1024_8_profile_compare.md
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
### Performance Comparison Report
|
| 2 |
+
|
| 3 |
+
#### 1. High-level Summary
|
| 4 |
+
| Metric | Baseline | New | Diff | Status |
|
| 5 |
+
| :--- | :--- | :--- | :--- | :--- |
|
| 6 |
+
| **E2E Latency** | 14858.56 ms | 28428.90 ms | **+13570.34 ms (+91.3%)** | ❌ |
|
| 7 |
+
| **Throughput** | 0.07 req/s | 0.04 req/s | - | - |
|
| 8 |
+
|
| 9 |
+
|
| 10 |
+
#### 2. Stage Breakdown
|
| 11 |
+
| Stage Name | Baseline (ms) | New (ms) | Diff (ms) | Diff (%) | Status |
|
| 12 |
+
| :--- | :--- | :--- | :--- | :--- | :--- |
|
| 13 |
+
| InputValidationStage | 1.79 | 0.83 | -0.96 | -53.8% | ⚪️ |
|
| 14 |
+
| TextEncodingStage | 602.41 | 620.45 | +18.04 | +3.0% | ⚪️ |
|
| 15 |
+
| LatentPreparationStage | 0.16 | 0.25 | +0.09 | +55.0% | ⚪️ |
|
| 16 |
+
| TimestepPreparationStage | 0.48 | 0.77 | +0.29 | +59.6% | ⚪️ |
|
| 17 |
+
| DenoisingStage | 14171.21 | 27742.68 | +13571.46 | +95.8% | 🔴 |
|
| 18 |
+
| DecodingStage | 56.27 | 10.83 | -45.45 | -80.8% | ⚪️ |
|
| 19 |
+
|
| 20 |
+
|
| 21 |
+
<details>
|
| 22 |
+
<summary>Metadata</summary>
|
| 23 |
+
|
| 24 |
+
- Baseline Commit: `d071bfde2540084b816b9b5c8f7396be0efb3cd4`
|
| 25 |
+
- New Commit: `c2b3e42ad64521c018443f7efe294658a0b1da3b`
|
| 26 |
+
- Timestamp: 2026-04-19T13:30:57.145199
|
| 27 |
+
</details>
|