fix

docs/source/en/concepts/paradigms_of_parallelism.md

@@ -140,3 +140,4 @@ Related paper:
 - [ZeRO-Offload: Democratizing Billion-Scale Model Training](https://arxiv.org/abs/2101.06840)
 - [ZeRO-Infinity: Breaking the GPU Memory Wall for Extreme Scale Deep Learning](https://arxiv.org/abs/2104.07857)
 - [PatrickStar: Parallel Training of Pre-trained Models via Chunk-based Memory Management](https://arxiv.org/abs/2108.05818)
+<!-- doc-test-command: echo  -->

docs/source/en/features/sequence_parallelism.md

@@ -126,7 +126,7 @@ plugin = HybridParallelPlugin(
         )
 ```
 #### Using Booster
-```
+```python
 booster = Booster(plugin=plugin)
 dataloader = plugin.prepare_dataloader(dataset, batch_size=args.batch_size, shuffle=True, drop_last=True, seed=42)
 model, optimizer, _, dataloader, _ = booster.boost(model, optimizer, dataloader=dataloader)

docs/source/zh-Hans/concepts/paradigms_of_parallelism.md

@@ -109,3 +109,4 @@ ring attention思路类似于flash attention，每个GPU只计算一个局部的
 - [ZeRO-Offload: Democratizing Billion-Scale Model Training](https://arxiv.org/abs/2101.06840)
 - [ZeRO-Infinity: Breaking the GPU Memory Wall for Extreme Scale Deep Learning](https://arxiv.org/abs/2104.07857)
 - [PatrickStar: Parallel Training of Pre-trained Models via Chunk-based Memory Management](https://arxiv.org/abs/2108.05818)
+<!-- doc-test-command: echo  -->

docs/source/zh-Hans/features/sequence_parallelism.md

@@ -125,7 +125,7 @@ plugin = HybridParallelPlugin(
         )
 ```
 #### 使用booster
-```
+```python
 booster = Booster(plugin=plugin)
 dataloader = plugin.prepare_dataloader(dataset, batch_size=args.batch_size, shuffle=True, drop_last=True, seed=42)
 model, optimizer, _, dataloader, _ = booster.boost(model, optimizer, dataloader=dataloader)