ZF/insightface
1
0
Fork 0
mirror of https://github.com/deepinsight/insightface.git synced 2026-05-20 00:10:28 +00:00
Code Issues Packages Projects Releases Wiki Activity

added resnet2060 training log

Browse Source
This commit is contained in:
anxiang
2021-06-23 04:00:08 +08:00
parent 43393dbb54
commit a2a3d74e09
1 changed files with 7 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
recognition/arcface_torch/README.md
View File

@@ -36,7 +36,7 @@ partial fc sampling strategy will get same accuracy with several times faster tr
1. Training speed of different parallel methods (samples/second), Tesla V100 32GB * 8. (Larger is better)
| Method | Bs128-R100-2 Million Identities | Bs128-R50-4 Million Identities | Bs64-R50-8 Million Identities |
| Method | Bs1024-R100-2 Million Identities | Bs1024-R50-4 Million Identities | Bs512-R50-8 Million Identities |
| :--- | :--- | :--- | :--- |
| Data Parallel | 1 | 1 | 1 |
| Model Parallel | 1362 | 1600 | 482 |
@@ -45,7 +45,7 @@ partial fc sampling strategy will get same accuracy with several times faster tr
2. GPU memory cost of different parallel methods (GB per GPU), Tesla V100 32GB * 8. (Smaller is better)
| Method | Bs128-R100-2 Million Identities | Bs128-R50-4 Million Identities | Bs64-R50-8 Million Identities |
| Method | Bs1024-R100-2 Million Identities | Bs1024-R50-4 Million Identities | Bs512-R50-8 Million Identities |
| :--- | :--- | :--- | :--- |
| Data Parallel | OOM | OOM | OOM |
| Model Parallel | 27.3 | 30.3 | 32.1 |
@@ -72,10 +72,11 @@ All Model Can be found in here.
### MS1MV3
| Datasets | log | backbone | IJBC(1e-05) | IJBC(1e-04) |agedb30|cfp_fp|lfw |
| :---: | :--- | :--- | :--- | :--- |:--- |:--- |:--- |
| MS1MV3-Arcface |[log](https://raw.githubusercontent.com/anxiangsir/insightface_arcface_log/master/ms1mv3_arcface_r18_fp16/training.log) | r18-fp16 | 92.07 | 94.66 | 97.77 | 97.73 | 99.77 |
| MS1MV3-Arcface |[log](https://raw.githubusercontent.com/anxiangsir/insightface_arcface_log/master/ms1mv3_arcface_r34_fp16/training.log) | r34-fp16 | 94.10 | 95.90 | 98.10 | 98.67 | 99.80 |
| MS1MV3-Arcface |[log](https://raw.githubusercontent.com/anxiangsir/insightface_arcface_log/master/ms1mv3_arcface_r50_fp16/training.log) | r50-fp16 | 94.79 | 96.46 | 98.35 | 98.96 | 99.83 |
| MS1MV3-Arcface |[log](https://raw.githubusercontent.com/anxiangsir/insightface_arcface_log/master/ms1mv3_arcface_r100_fp16/training.log) | r100-fp16 | 95.31 | 96.81 | 98.48 | 99.06 | 99.85 |
| MS1MV3-Arcface |[log](https://raw.githubusercontent.com/anxiangsir/insightface_arcface_log/master/ms1mv3_arcface_r18_fp16/training.log) | r18-fp16 | 92.07 | 94.66 | 97.77 | 97.73 | 99.77 |
| MS1MV3-Arcface |[log](https://raw.githubusercontent.com/anxiangsir/insightface_arcface_log/master/ms1mv3_arcface_r34_fp16/training.log) | r34-fp16 | 94.10 | 95.90 | 98.10 | 98.67 | 99.80 |
| MS1MV3-Arcface |[log](https://raw.githubusercontent.com/anxiangsir/insightface_arcface_log/master/ms1mv3_arcface_r50_fp16/training.log) | r50-fp16 | 94.79 | 96.46 | 98.35 | 98.96 | 99.83 |
| MS1MV3-Arcface |[log](https://raw.githubusercontent.com/anxiangsir/insightface_arcface_log/master/ms1mv3_arcface_r100_fp16/training.log) | r100-fp16 | 95.31 | 96.81 | 98.48 | 99.06 | 99.85 |
| MS1MV3-Arcface |[log](https://raw.githubusercontent.com/anxiangsir/insightface_arcface_log/master/ms1mv3_arcface_r2060_fp16/training.log)| **r2060-fp16**| 95.34 | 97.11 | 98.67 | 99.24 | 99.87 |
### Glint360k
| Datasets | log |backbone | IJBC(1e-05) | IJBC(1e-04) |agedb30|cfp_fp|lfw |