fix Torch's docs

tp-nan · tp-nan · commit 23816afb7035 · 2023-11-06T10:30:54.000+08:00
diff --git a/docs/backend-reference/torch.mdx b/docs/backend-reference/torch.mdx
@@ -94,7 +94,7 @@ flowchart LR
 
 ## Torch
 Similar to [SyncTensor](#synctensor), but with additional cross-card functionality. Effective from version 0.3.2b1.
-- Torch::backend: required
+- Torch::backend: required, like `Torch[TensorrtTensor]`
 - device_id: Default is -1, which sets the current device to this ID. During initialization, it is equivalent to calling `c10::cuda::set_device(device_id)` or `torch.cuda.set_device(device_id)`. During forward propagation, the input data type must be `at::Tensor` or `vector<at::Tensor>`. This backend will move the data to the specified graphics card.
 
 ## SaveTensor
diff --git a/i18n/zh/docusaurus-plugin-content-docs/current/backend-reference/torch.mdx b/i18n/zh/docusaurus-plugin-content-docs/current/backend-reference/torch.mdx
@@ -97,7 +97,7 @@ flowchart LR
 
 ## Torch
 &#19982;[SyncTensor](#synctensor)&#31867;&#20284;&#65292;&#20294;&#39069;&#22806;&#28155;&#21152;&#20102;&#36328;&#21345;&#21151;&#33021;&#12290; &#20174;0.3.2b1&#29256;&#26412;&#24320;&#22987;&#29983;&#25928;&#12290;
-- Torch::backend: default=Identity
+- Torch::backend: &#24517;&#39035;&#25552;&#20379;, &#27604;&#22914; `Torch[TensorrtTensor]`
 - device_id: &#40664;&#35748;&#20026;-1&#65307;&#34920;&#31034;&#23558;&#24403;&#21069;&#35774;&#22791;&#35774;&#32622;&#20026;&#36825;&#20010;id. &#21021;&#22987;&#21270;&#26102;&#20869;&#37096;&#30456;&#24403;&#20110;&#35843;&#29992;&#20102;`c10::cuda::set_device(device_id)` &#25110;&#32773; `torch.cuda.set_device(device_id)`&#65292;&#21069;&#21521;&#26102;&#65292;&#35201;&#27714;&#36755;&#20837;&#25968;&#25454;&#31867;&#22411;&#26159;`at::Tensor` or `vector<at::Tensor>`; &#27492;&#21518;&#31471;&#20250;&#23558;&#25968;&#25454;&#25644;&#36816;&#21040;&#25351;&#23450;&#26174;&#21345;&#19978;&#12290;
 
 ## SaveTensor

-Original file line number
+Diff line change
 ## Torch
 与[SyncTensor](#synctensor)类似，但额外添加了跨卡功能。 从0.3.2b1版本开始生效。
 -- Torch::backend: default=Identity
 +- Torch::backend: 必须提供, 比如 `Torch[TensorrtTensor]`
 - device_id: 默认为-1；表示将当前设备设置为这个id. 初始化时内部相当于调用了`c10::cuda::set_device(device_id)` 或者 `torch.cuda.set_device(device_id)`，前向时，要求输入数据类型是`at::Tensor` or `vector<at::Tensor>`; 此后端会将数据搬运到指定显卡上。
 ## SaveTensor