Skip to content
Permalink

Comparing changes

This is a direct comparison between two commits made in this repository or its related repositories. View the default comparison for this range or learn more about diff comparisons.

Open a pull request

Create a new pull request by comparing changes across two branches. If you need to, you can also . Learn more about diff comparisons here.
base repository: deepinsight/insightface
Failed to load repositories. Confirm that selected base ref is valid, then try again.
Loading
base: 15bda1666c87db71be8bbcfddf2143b3279066ac
Choose a base ref
..
head repository: deepinsight/insightface
Failed to load repositories. Confirm that selected head ref is valid, then try again.
Loading
compare: 88b0739fc26344001f23f27441ed88a0999b5212
Choose a head ref
Showing with 4 additions and 4 deletions.
  1. +4 −4 recognition/arcface_torch/README.md
8 changes: 4 additions & 4 deletions recognition/arcface_torch/README.md
Original file line number Diff line number Diff line change
@@ -52,18 +52,18 @@ ArcFace_torch can train large-scale face recognition training set efficiently an
When the number of classes in training sets is greater than 300K and the training is sufficient,
partial fc sampling strategy will get same accuracy with several times faster training performance and smaller GPU memory.

1. Different Parallel Methods Training Speed
1. Different Parallel Methods Training Speed (Larger is better)

| Method | Bs128-r100-2 Million Identities | Bs128-r50-4 Million Identities | Bs64-r50-8 Million Identities |
| Method | Bs128-R100-2 Million Identities | Bs128-R50-4 Million Identities | Bs64-R50-8 Million Identities |
| :--- | :--- | :--- | :--- |
| Data Parallel | 1 | 1 | 1 |
| Model Parallel | 1362 | 1600 | 482 |
| Fp16 + Model Parallel | 2006 | 2165 | 767 |
| Fp16 + Partial Fc 0.1 | 3247 | 4385 | 3001 |

2. Different Parallel Methods GPU Memory
2. Different Parallel Methods GPU Memory (Smaller is better)

| Method | Bs128-r100-2 Million Identities | Bs128-r50-4 Million Identities | Bs64-r50-8 Million Identities |
| Method | Bs128-R100-2 Million Identities | Bs128-R50-4 Million Identities | Bs64-R50-8 Million Identities |
| :--- | :--- | :--- | :--- |
| Data Parallel | OOM | OOM | OOM |
| Model Parallel | 27382 | 30322 | 32182 |