@@ -47,12 +47,14 @@ Increasing this value generally increases the throughput.
47
47
If this setting is greater than the number of hardware threads
48
48
it will automatically be changed to a value less than the number of hardware threads.
49
49
Defaults to 1.
50
-
50
+ +
51
+ --
51
52
[NOTE]
52
53
=============================================
53
- If the sum of `threads_per_allocation` and `number_of_allocations` is greater than the number of
54
- hardware threads then the number of `inference_threads` will be reduced.
54
+ If the sum of `threads_per_allocation` and `number_of_allocations` is greater
55
+ than the number of hardware threads, the `threads_per_allocation` value is reduced.
55
56
=============================================
57
+ --
56
58
57
59
`queue_capacity`::
58
60
(Optional, integer)
@@ -103,7 +105,10 @@ The API returns the following results:
103
105
"assignment": {
104
106
"task_parameters": {
105
107
"model_id": "elastic__distilbert-base-uncased-finetuned-conll03-english",
106
- "model_bytes": 265632637
108
+ "model_bytes": 265632637,
109
+ "threads_per_allocation" : 1,
110
+ "number_of_allocations" : 1,
111
+ "queue_capacity" : 1024
107
112
},
108
113
"routing_table": {
109
114
"uckeG3R8TLe2MMNBQ6AGrw": {
@@ -112,7 +117,7 @@ The API returns the following results:
112
117
}
113
118
},
114
119
"assignment_state": "started",
115
- "start_time": "2021 -11-02T11:50:34.766591Z"
120
+ "start_time": "2022 -11-02T11:50:34.766591Z"
116
121
}
117
122
}
118
123
----
0 commit comments