iicd/xautodl
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update code styles

Browse Source
This commit is contained in:
D-X-Y
2020-01-09 22:26:23 +11:00
parent 5ac5060a33
commit ad34af9913
26 changed files with 192 additions and 81 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
NAS-Bench-102.md
View File

@@ -6,9 +6,9 @@ Each edge here is associated with an operation selected from a predefined operat
For it to be applicable for all NAS algorithms, the search space defined in NAS-Bench-102 includes 4 nodes and 5 associated operation options, which generates 15,625 neural cell candidates in total.
In this Markdown file, we provide:
- [How to Use NAS-Bench-102](#how-to-use-nas-bench-102)
- [Instruction to re-generate NAS-Bench-102](#instruction-to-re-generate-nas-bench-102)
- [10 NAS algorithms evaluated in our paper](#to-reproduce-10-baseline-nas-algorithms-in-nas-bench-102)
- [How to Use NAS-Bench-102](#how-to-use-nas-bench-102)
- [Instruction to re-generate NAS-Bench-102](#instruction-to-re-generate-nas-bench-102)
- [10 NAS algorithms evaluated in our paper](#to-reproduce-10-baseline-nas-algorithms-in-nas-bench-102)
Note: please use `PyTorch >= 1.2.0` and `Python >= 3.6.0`.
@@ -140,6 +140,8 @@ This command will train 390 architectures (id from 0 to 389) using the following
| CIFAR-100 | train | valid / test |
| ImageNet-16-120 | train | valid / test |
Note that the above `train`, `valid`, and `test` indicate the proposed splits in our NAS-Bench-102, and they might be different with the original splits.
3. calculate the latency, merge the results of all architectures, and simplify the results.
(see commands in `output/NAS-BENCH-102-4/meta-node-4.cal-script.txt` which is automatically generated by step-1).
```
@@ -167,7 +169,7 @@ If researchers can provide better results with different hyper-parameters, we ar
**Note that** you need to prepare the training and test data as described in [Preparation and Download](#preparation-and-download)
- [1] `CUDA_VISIBLE_DEVICES=0 bash ./scripts-search/algos/DARTS-V1.sh cifar10 -1`
- [1] `CUDA_VISIBLE_DEVICES=0 bash ./scripts-search/algos/DARTS-V1.sh cifar10 -1`, where `cifar10` can be replaced with `cifar100` or `ImageNet16-120`.
- [2] `CUDA_VISIBLE_DEVICES=0 bash ./scripts-search/algos/DARTS-V2.sh cifar10 -1`
- [3] `CUDA_VISIBLE_DEVICES=0 bash ./scripts-search/algos/GDAS.sh cifar10 -1`
- [4] `CUDA_VISIBLE_DEVICES=0 bash ./scripts-search/algos/SETN.sh cifar10 -1`