diff --git a/docs/user_guide/model_configuration.md b/docs/user_guide/model_configuration.md index 241301ade7..023fcf259b 100644 --- a/docs/user_guide/model_configuration.md +++ b/docs/user_guide/model_configuration.md @@ -723,7 +723,7 @@ The above configuration creates 3 model instances, one on each device themselves as "R1" is local for their own device, however, they will contend for "R2" because it is specified as a global resource which means "R2" is shared across the system. Though these instances don't -contend for "R1" among themsleves, but they will contend for "R1" +contend for "R1" among themselves, but they will contend for "R1" with other model instances which includes "R1" in their resource requirements and run on the same device as them. diff --git a/docs/user_guide/response_cache.md b/docs/user_guide/response_cache.md index e70085e798..ebeef4cb0a 100644 --- a/docs/user_guide/response_cache.md +++ b/docs/user_guide/response_cache.md @@ -58,16 +58,20 @@ sections below for more details. ### Enable Caching on Server-side -The response cache is enabled on the server-side by specifying a -`` and corresponding configuration when starting +The response cache is enabled on the server-side by specifying a cache +implementation name `` and corresponding configuration when starting the Triton server. Through the CLI, this translates to setting -`tritonserver --cache-config ,= ...`. For example: +`tritonserver --cache-config ,= ...`. For example: ``` tritonserver --cache-config local,size=1048576 ``` +> [!NOTE] +> If using a non-interactive shell, you may need to specify the argument without +> the space like so: `--cache-config=,=`. + For in-process C API applications, this translates to calling `TRITONSERVER_SetCacheConfig(const char* cache_implementation, const char* config_json)`. diff --git a/docs/user_guide/trace.md b/docs/user_guide/trace.md index 1abe019ea1..d359299499 100644 --- a/docs/user_guide/trace.md +++ b/docs/user_guide/trace.md @@ -543,7 +543,7 @@ The following table shows available OpenTelemetry trace APIs settings for For example:
--trace-config opentelemetry,resource=service.name=triton
--trace-config opentelemetry,resource=service.version=1
- Alternatively, key-vaue attributes can be specified through
+ Alternatively, key-value attributes can be specified through
OTEL_RESOURCE_ATTRIBUTES environment variable.