gpustack-runner 0.1.24__py3-none-any.whl → 0.1.24.post2__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- gpustack_runner/_version.py +2 -2
- gpustack_runner/_version_appendix.py +1 -1
- gpustack_runner/cmds/images.py +2 -11
- gpustack_runner/runner.py +0 -6
- gpustack_runner/runner.py.json +220 -44
- {gpustack_runner-0.1.24.dist-info → gpustack_runner-0.1.24.post2.dist-info}/METADATA +30 -44
- gpustack_runner-0.1.24.post2.dist-info/RECORD +17 -0
- gpustack_runner-0.1.24.dist-info/RECORD +0 -17
- {gpustack_runner-0.1.24.dist-info → gpustack_runner-0.1.24.post2.dist-info}/WHEEL +0 -0
- {gpustack_runner-0.1.24.dist-info → gpustack_runner-0.1.24.post2.dist-info}/entry_points.txt +0 -0
- {gpustack_runner-0.1.24.dist-info → gpustack_runner-0.1.24.post2.dist-info}/licenses/LICENSE +0 -0
gpustack_runner/_version.py
CHANGED
|
@@ -27,8 +27,8 @@ version_tuple: VERSION_TUPLE
|
|
|
27
27
|
__commit_id__: COMMIT_ID
|
|
28
28
|
commit_id: COMMIT_ID
|
|
29
29
|
|
|
30
|
-
__version__ = version = '0.1.24'
|
|
31
|
-
__version_tuple__ = version_tuple = (0, 1, 24)
|
|
30
|
+
__version__ = version = '0.1.24.post2'
|
|
31
|
+
__version_tuple__ = version_tuple = (0, 1, 24, 'post2')
|
|
32
32
|
try:
|
|
33
33
|
from ._version_appendix import git_commit
|
|
34
34
|
__commit_id__ = commit_id = git_commit
|
|
@@ -1 +1 @@
|
|
|
1
|
-
git_commit = "
|
|
1
|
+
git_commit = "62d75c6"
|
gpustack_runner/cmds/images.py
CHANGED
|
@@ -464,7 +464,7 @@ class SaveImagesSubCommand(SubCommand):
|
|
|
464
464
|
# Submit tasks
|
|
465
465
|
for task_name, src_img, dst_file, dst_file_relative in saving_tasks:
|
|
466
466
|
if dst_file.exists():
|
|
467
|
-
print(f"{dst_file
|
|
467
|
+
print(f"✅ {dst_file} already exists, skipping save {src_img}.")
|
|
468
468
|
continue
|
|
469
469
|
dst_file.parent.mkdir(parents=True, exist_ok=True, mode=0o744)
|
|
470
470
|
|
|
@@ -976,7 +976,7 @@ class LoadImagesSubCommand(SubCommand):
|
|
|
976
976
|
load_parser.add_argument(
|
|
977
977
|
"input",
|
|
978
978
|
nargs=OPTIONAL,
|
|
979
|
-
help="Input directory to load images (default: current working directory)",
|
|
979
|
+
help="Input directory to load images from (default: current working directory)",
|
|
980
980
|
)
|
|
981
981
|
|
|
982
982
|
load_parser.set_defaults(func=LoadImagesSubCommand)
|
|
@@ -1348,10 +1348,6 @@ def list_images(**kwargs) -> list[PlatformedImage]:
|
|
|
1348
1348
|
A list of platformed images.
|
|
1349
1349
|
|
|
1350
1350
|
"""
|
|
1351
|
-
# Reset to default for listing images,
|
|
1352
|
-
# in case the env is set to other value.
|
|
1353
|
-
envs.GPUSTACK_RUNNER_DEFAULT_CONTAINER_NAMESPACE = None
|
|
1354
|
-
|
|
1355
1351
|
platform = kwargs.pop("platform", None)
|
|
1356
1352
|
repository = kwargs.pop("repository", None)
|
|
1357
1353
|
|
|
@@ -1497,8 +1493,3 @@ def _execute_command(
|
|
|
1497
1493
|
args=command,
|
|
1498
1494
|
returncode=returncode,
|
|
1499
1495
|
)
|
|
1500
|
-
|
|
1501
|
-
|
|
1502
|
-
append_images(
|
|
1503
|
-
"gpustack/runtime:pause",
|
|
1504
|
-
)
|
gpustack_runner/runner.py
CHANGED
|
@@ -10,8 +10,6 @@ from typing import Any
|
|
|
10
10
|
|
|
11
11
|
from dataclasses_json import dataclass_json
|
|
12
12
|
|
|
13
|
-
from . import envs
|
|
14
|
-
|
|
15
13
|
_RE_DOCKER_IMAGE = re.compile(
|
|
16
14
|
r"(?:(?P<prefix>[\w\\.\-]+(?:/[\w\\.\-]+)*)/)?runner:(?P<backend>(Host|cann|corex|cuda|dtk|hggc|maca|musa|rocm))(?P<backend_version>[XY\d\\.]+)(?:-(?P<backend_variant>\w+))?-(?P<service>(vllm|voxbox|mindie|sglang))(?P<service_version>[\w\\.]+)(?:-(?P<suffix>\w+))?",
|
|
17
15
|
)
|
|
@@ -239,10 +237,6 @@ def list_runners(**kwargs) -> Runners | list[dict]:
|
|
|
239
237
|
json_list = json.load(f)
|
|
240
238
|
runners = []
|
|
241
239
|
for item in json_list:
|
|
242
|
-
if namespace := envs.GPUSTACK_RUNNER_DEFAULT_CONTAINER_NAMESPACE:
|
|
243
|
-
docker_image = item["docker_image"]
|
|
244
|
-
docker_image = docker_image.replace("gpustack/", f"{namespace}/")
|
|
245
|
-
item["docker_image"] = docker_image
|
|
246
240
|
runners.append(Runner.from_dict(item))
|
|
247
241
|
|
|
248
242
|
todict = kwargs.pop("todict", False)
|
gpustack_runner/runner.py.json
CHANGED
|
@@ -1,35 +1,79 @@
|
|
|
1
1
|
[
|
|
2
2
|
{
|
|
3
3
|
"backend": "cann",
|
|
4
|
-
"backend_version": "8.
|
|
5
|
-
"original_backend_version": "8.
|
|
4
|
+
"backend_version": "8.5",
|
|
5
|
+
"original_backend_version": "8.5.0",
|
|
6
6
|
"backend_variant": "a3",
|
|
7
7
|
"service": "mindie",
|
|
8
|
-
"service_version": "2.
|
|
8
|
+
"service_version": "2.3.0",
|
|
9
9
|
"platform": "linux/amd64",
|
|
10
|
-
"docker_image": "gpustack/runner:cann8.
|
|
10
|
+
"docker_image": "gpustack/runner:cann8.5-a3-mindie2.3.0",
|
|
11
11
|
"deprecated": false
|
|
12
12
|
},
|
|
13
13
|
{
|
|
14
14
|
"backend": "cann",
|
|
15
|
-
"backend_version": "8.
|
|
16
|
-
"original_backend_version": "8.
|
|
15
|
+
"backend_version": "8.5",
|
|
16
|
+
"original_backend_version": "8.5.0",
|
|
17
17
|
"backend_variant": "a3",
|
|
18
18
|
"service": "mindie",
|
|
19
|
-
"service_version": "2.
|
|
19
|
+
"service_version": "2.3.0",
|
|
20
20
|
"platform": "linux/arm64",
|
|
21
|
-
"docker_image": "gpustack/runner:cann8.
|
|
21
|
+
"docker_image": "gpustack/runner:cann8.5-a3-mindie2.3.0",
|
|
22
22
|
"deprecated": false
|
|
23
23
|
},
|
|
24
24
|
{
|
|
25
25
|
"backend": "cann",
|
|
26
|
-
"backend_version": "8.
|
|
27
|
-
"original_backend_version": "8.
|
|
26
|
+
"backend_version": "8.5",
|
|
27
|
+
"original_backend_version": "8.5.0",
|
|
28
28
|
"backend_variant": "a3",
|
|
29
29
|
"service": "sglang",
|
|
30
|
-
"service_version": "0.5.
|
|
30
|
+
"service_version": "0.5.8",
|
|
31
31
|
"platform": "linux/arm64",
|
|
32
|
-
"docker_image": "gpustack/runner:cann8.
|
|
32
|
+
"docker_image": "gpustack/runner:cann8.5-a3-sglang0.5.8",
|
|
33
|
+
"deprecated": false
|
|
34
|
+
},
|
|
35
|
+
{
|
|
36
|
+
"backend": "cann",
|
|
37
|
+
"backend_version": "8.5",
|
|
38
|
+
"original_backend_version": "8.5.0",
|
|
39
|
+
"backend_variant": "a3",
|
|
40
|
+
"service": "vllm",
|
|
41
|
+
"service_version": "0.14.1",
|
|
42
|
+
"platform": "linux/amd64",
|
|
43
|
+
"docker_image": "gpustack/runner:cann8.5-a3-vllm0.14.1",
|
|
44
|
+
"deprecated": false
|
|
45
|
+
},
|
|
46
|
+
{
|
|
47
|
+
"backend": "cann",
|
|
48
|
+
"backend_version": "8.5",
|
|
49
|
+
"original_backend_version": "8.5.0",
|
|
50
|
+
"backend_variant": "a3",
|
|
51
|
+
"service": "vllm",
|
|
52
|
+
"service_version": "0.14.1",
|
|
53
|
+
"platform": "linux/arm64",
|
|
54
|
+
"docker_image": "gpustack/runner:cann8.5-a3-vllm0.14.1",
|
|
55
|
+
"deprecated": false
|
|
56
|
+
},
|
|
57
|
+
{
|
|
58
|
+
"backend": "cann",
|
|
59
|
+
"backend_version": "8.5",
|
|
60
|
+
"original_backend_version": "8.5.0",
|
|
61
|
+
"backend_variant": "a3",
|
|
62
|
+
"service": "vllm",
|
|
63
|
+
"service_version": "0.13.0",
|
|
64
|
+
"platform": "linux/amd64",
|
|
65
|
+
"docker_image": "gpustack/runner:cann8.5-a3-vllm0.13.0",
|
|
66
|
+
"deprecated": false
|
|
67
|
+
},
|
|
68
|
+
{
|
|
69
|
+
"backend": "cann",
|
|
70
|
+
"backend_version": "8.5",
|
|
71
|
+
"original_backend_version": "8.5.0",
|
|
72
|
+
"backend_variant": "a3",
|
|
73
|
+
"service": "vllm",
|
|
74
|
+
"service_version": "0.13.0",
|
|
75
|
+
"platform": "linux/arm64",
|
|
76
|
+
"docker_image": "gpustack/runner:cann8.5-a3-vllm0.13.0",
|
|
33
77
|
"deprecated": false
|
|
34
78
|
},
|
|
35
79
|
{
|
|
@@ -37,10 +81,21 @@
|
|
|
37
81
|
"backend_version": "8.3",
|
|
38
82
|
"original_backend_version": "8.3.rc2",
|
|
39
83
|
"backend_variant": "a3",
|
|
40
|
-
"service": "
|
|
41
|
-
"service_version": "
|
|
84
|
+
"service": "mindie",
|
|
85
|
+
"service_version": "2.2.rc1",
|
|
86
|
+
"platform": "linux/amd64",
|
|
87
|
+
"docker_image": "gpustack/runner:cann8.3-a3-mindie2.2.rc1",
|
|
88
|
+
"deprecated": false
|
|
89
|
+
},
|
|
90
|
+
{
|
|
91
|
+
"backend": "cann",
|
|
92
|
+
"backend_version": "8.3",
|
|
93
|
+
"original_backend_version": "8.3.rc2",
|
|
94
|
+
"backend_variant": "a3",
|
|
95
|
+
"service": "mindie",
|
|
96
|
+
"service_version": "2.2.rc1",
|
|
42
97
|
"platform": "linux/arm64",
|
|
43
|
-
"docker_image": "gpustack/runner:cann8.3-a3-
|
|
98
|
+
"docker_image": "gpustack/runner:cann8.3-a3-mindie2.2.rc1",
|
|
44
99
|
"deprecated": false
|
|
45
100
|
},
|
|
46
101
|
{
|
|
@@ -48,10 +103,10 @@
|
|
|
48
103
|
"backend_version": "8.3",
|
|
49
104
|
"original_backend_version": "8.3.rc2",
|
|
50
105
|
"backend_variant": "a3",
|
|
51
|
-
"service": "
|
|
52
|
-
"service_version": "0.
|
|
53
|
-
"platform": "linux/
|
|
54
|
-
"docker_image": "gpustack/runner:cann8.3-a3-
|
|
106
|
+
"service": "sglang",
|
|
107
|
+
"service_version": "0.5.7",
|
|
108
|
+
"platform": "linux/arm64",
|
|
109
|
+
"docker_image": "gpustack/runner:cann8.3-a3-sglang0.5.7",
|
|
55
110
|
"deprecated": false
|
|
56
111
|
},
|
|
57
112
|
{
|
|
@@ -59,10 +114,10 @@
|
|
|
59
114
|
"backend_version": "8.3",
|
|
60
115
|
"original_backend_version": "8.3.rc2",
|
|
61
116
|
"backend_variant": "a3",
|
|
62
|
-
"service": "
|
|
63
|
-
"service_version": "0.
|
|
117
|
+
"service": "sglang",
|
|
118
|
+
"service_version": "0.5.6.post2",
|
|
64
119
|
"platform": "linux/arm64",
|
|
65
|
-
"docker_image": "gpustack/runner:cann8.3-a3-
|
|
120
|
+
"docker_image": "gpustack/runner:cann8.3-a3-sglang0.5.6.post2",
|
|
66
121
|
"deprecated": false
|
|
67
122
|
},
|
|
68
123
|
{
|
|
@@ -221,35 +276,79 @@
|
|
|
221
276
|
},
|
|
222
277
|
{
|
|
223
278
|
"backend": "cann",
|
|
224
|
-
"backend_version": "8.
|
|
225
|
-
"original_backend_version": "8.
|
|
279
|
+
"backend_version": "8.5",
|
|
280
|
+
"original_backend_version": "8.5.0",
|
|
226
281
|
"backend_variant": "910b",
|
|
227
282
|
"service": "mindie",
|
|
228
|
-
"service_version": "2.
|
|
283
|
+
"service_version": "2.3.0",
|
|
229
284
|
"platform": "linux/amd64",
|
|
230
|
-
"docker_image": "gpustack/runner:cann8.
|
|
285
|
+
"docker_image": "gpustack/runner:cann8.5-910b-mindie2.3.0",
|
|
231
286
|
"deprecated": false
|
|
232
287
|
},
|
|
233
288
|
{
|
|
234
289
|
"backend": "cann",
|
|
235
|
-
"backend_version": "8.
|
|
236
|
-
"original_backend_version": "8.
|
|
290
|
+
"backend_version": "8.5",
|
|
291
|
+
"original_backend_version": "8.5.0",
|
|
237
292
|
"backend_variant": "910b",
|
|
238
293
|
"service": "mindie",
|
|
239
|
-
"service_version": "2.
|
|
294
|
+
"service_version": "2.3.0",
|
|
240
295
|
"platform": "linux/arm64",
|
|
241
|
-
"docker_image": "gpustack/runner:cann8.
|
|
296
|
+
"docker_image": "gpustack/runner:cann8.5-910b-mindie2.3.0",
|
|
242
297
|
"deprecated": false
|
|
243
298
|
},
|
|
244
299
|
{
|
|
245
300
|
"backend": "cann",
|
|
246
|
-
"backend_version": "8.
|
|
247
|
-
"original_backend_version": "8.
|
|
301
|
+
"backend_version": "8.5",
|
|
302
|
+
"original_backend_version": "8.5.0",
|
|
248
303
|
"backend_variant": "910b",
|
|
249
304
|
"service": "sglang",
|
|
250
|
-
"service_version": "0.5.
|
|
305
|
+
"service_version": "0.5.8",
|
|
251
306
|
"platform": "linux/arm64",
|
|
252
|
-
"docker_image": "gpustack/runner:cann8.
|
|
307
|
+
"docker_image": "gpustack/runner:cann8.5-910b-sglang0.5.8",
|
|
308
|
+
"deprecated": false
|
|
309
|
+
},
|
|
310
|
+
{
|
|
311
|
+
"backend": "cann",
|
|
312
|
+
"backend_version": "8.5",
|
|
313
|
+
"original_backend_version": "8.5.0",
|
|
314
|
+
"backend_variant": "910b",
|
|
315
|
+
"service": "vllm",
|
|
316
|
+
"service_version": "0.14.1",
|
|
317
|
+
"platform": "linux/amd64",
|
|
318
|
+
"docker_image": "gpustack/runner:cann8.5-910b-vllm0.14.1",
|
|
319
|
+
"deprecated": false
|
|
320
|
+
},
|
|
321
|
+
{
|
|
322
|
+
"backend": "cann",
|
|
323
|
+
"backend_version": "8.5",
|
|
324
|
+
"original_backend_version": "8.5.0",
|
|
325
|
+
"backend_variant": "910b",
|
|
326
|
+
"service": "vllm",
|
|
327
|
+
"service_version": "0.14.1",
|
|
328
|
+
"platform": "linux/arm64",
|
|
329
|
+
"docker_image": "gpustack/runner:cann8.5-910b-vllm0.14.1",
|
|
330
|
+
"deprecated": false
|
|
331
|
+
},
|
|
332
|
+
{
|
|
333
|
+
"backend": "cann",
|
|
334
|
+
"backend_version": "8.5",
|
|
335
|
+
"original_backend_version": "8.5.0",
|
|
336
|
+
"backend_variant": "910b",
|
|
337
|
+
"service": "vllm",
|
|
338
|
+
"service_version": "0.13.0",
|
|
339
|
+
"platform": "linux/amd64",
|
|
340
|
+
"docker_image": "gpustack/runner:cann8.5-910b-vllm0.13.0",
|
|
341
|
+
"deprecated": false
|
|
342
|
+
},
|
|
343
|
+
{
|
|
344
|
+
"backend": "cann",
|
|
345
|
+
"backend_version": "8.5",
|
|
346
|
+
"original_backend_version": "8.5.0",
|
|
347
|
+
"backend_variant": "910b",
|
|
348
|
+
"service": "vllm",
|
|
349
|
+
"service_version": "0.13.0",
|
|
350
|
+
"platform": "linux/arm64",
|
|
351
|
+
"docker_image": "gpustack/runner:cann8.5-910b-vllm0.13.0",
|
|
253
352
|
"deprecated": false
|
|
254
353
|
},
|
|
255
354
|
{
|
|
@@ -257,10 +356,21 @@
|
|
|
257
356
|
"backend_version": "8.3",
|
|
258
357
|
"original_backend_version": "8.3.rc2",
|
|
259
358
|
"backend_variant": "910b",
|
|
260
|
-
"service": "
|
|
261
|
-
"service_version": "
|
|
359
|
+
"service": "mindie",
|
|
360
|
+
"service_version": "2.2.rc1",
|
|
361
|
+
"platform": "linux/amd64",
|
|
362
|
+
"docker_image": "gpustack/runner:cann8.3-910b-mindie2.2.rc1",
|
|
363
|
+
"deprecated": false
|
|
364
|
+
},
|
|
365
|
+
{
|
|
366
|
+
"backend": "cann",
|
|
367
|
+
"backend_version": "8.3",
|
|
368
|
+
"original_backend_version": "8.3.rc2",
|
|
369
|
+
"backend_variant": "910b",
|
|
370
|
+
"service": "mindie",
|
|
371
|
+
"service_version": "2.2.rc1",
|
|
262
372
|
"platform": "linux/arm64",
|
|
263
|
-
"docker_image": "gpustack/runner:cann8.3-910b-
|
|
373
|
+
"docker_image": "gpustack/runner:cann8.3-910b-mindie2.2.rc1",
|
|
264
374
|
"deprecated": false
|
|
265
375
|
},
|
|
266
376
|
{
|
|
@@ -268,10 +378,10 @@
|
|
|
268
378
|
"backend_version": "8.3",
|
|
269
379
|
"original_backend_version": "8.3.rc2",
|
|
270
380
|
"backend_variant": "910b",
|
|
271
|
-
"service": "
|
|
272
|
-
"service_version": "0.
|
|
273
|
-
"platform": "linux/
|
|
274
|
-
"docker_image": "gpustack/runner:cann8.3-910b-
|
|
381
|
+
"service": "sglang",
|
|
382
|
+
"service_version": "0.5.7",
|
|
383
|
+
"platform": "linux/arm64",
|
|
384
|
+
"docker_image": "gpustack/runner:cann8.3-910b-sglang0.5.7",
|
|
275
385
|
"deprecated": false
|
|
276
386
|
},
|
|
277
387
|
{
|
|
@@ -279,10 +389,10 @@
|
|
|
279
389
|
"backend_version": "8.3",
|
|
280
390
|
"original_backend_version": "8.3.rc2",
|
|
281
391
|
"backend_variant": "910b",
|
|
282
|
-
"service": "
|
|
283
|
-
"service_version": "0.
|
|
392
|
+
"service": "sglang",
|
|
393
|
+
"service_version": "0.5.6.post2",
|
|
284
394
|
"platform": "linux/arm64",
|
|
285
|
-
"docker_image": "gpustack/runner:cann8.3-910b-
|
|
395
|
+
"docker_image": "gpustack/runner:cann8.3-910b-sglang0.5.6.post2",
|
|
286
396
|
"deprecated": false
|
|
287
397
|
},
|
|
288
398
|
{
|
|
@@ -527,6 +637,50 @@
|
|
|
527
637
|
"docker_image": "gpustack/runner:cann8.2-910b-vllm0.9.1",
|
|
528
638
|
"deprecated": false
|
|
529
639
|
},
|
|
640
|
+
{
|
|
641
|
+
"backend": "cann",
|
|
642
|
+
"backend_version": "8.5",
|
|
643
|
+
"original_backend_version": "8.5.0",
|
|
644
|
+
"backend_variant": "310p",
|
|
645
|
+
"service": "mindie",
|
|
646
|
+
"service_version": "2.3.0",
|
|
647
|
+
"platform": "linux/amd64",
|
|
648
|
+
"docker_image": "gpustack/runner:cann8.5-310p-mindie2.3.0",
|
|
649
|
+
"deprecated": false
|
|
650
|
+
},
|
|
651
|
+
{
|
|
652
|
+
"backend": "cann",
|
|
653
|
+
"backend_version": "8.5",
|
|
654
|
+
"original_backend_version": "8.5.0",
|
|
655
|
+
"backend_variant": "310p",
|
|
656
|
+
"service": "mindie",
|
|
657
|
+
"service_version": "2.3.0",
|
|
658
|
+
"platform": "linux/arm64",
|
|
659
|
+
"docker_image": "gpustack/runner:cann8.5-310p-mindie2.3.0",
|
|
660
|
+
"deprecated": false
|
|
661
|
+
},
|
|
662
|
+
{
|
|
663
|
+
"backend": "cann",
|
|
664
|
+
"backend_version": "8.5",
|
|
665
|
+
"original_backend_version": "8.5.0",
|
|
666
|
+
"backend_variant": "310p",
|
|
667
|
+
"service": "vllm",
|
|
668
|
+
"service_version": "0.14.1",
|
|
669
|
+
"platform": "linux/amd64",
|
|
670
|
+
"docker_image": "gpustack/runner:cann8.5-310p-vllm0.14.1",
|
|
671
|
+
"deprecated": false
|
|
672
|
+
},
|
|
673
|
+
{
|
|
674
|
+
"backend": "cann",
|
|
675
|
+
"backend_version": "8.5",
|
|
676
|
+
"original_backend_version": "8.5.0",
|
|
677
|
+
"backend_variant": "310p",
|
|
678
|
+
"service": "vllm",
|
|
679
|
+
"service_version": "0.14.1",
|
|
680
|
+
"platform": "linux/arm64",
|
|
681
|
+
"docker_image": "gpustack/runner:cann8.5-310p-vllm0.14.1",
|
|
682
|
+
"deprecated": false
|
|
683
|
+
},
|
|
530
684
|
{
|
|
531
685
|
"backend": "cann",
|
|
532
686
|
"backend_version": "8.3",
|
|
@@ -1418,6 +1572,28 @@
|
|
|
1418
1572
|
"docker_image": "gpustack/runner:hggc12.3-vllm0.11.1",
|
|
1419
1573
|
"deprecated": false
|
|
1420
1574
|
},
|
|
1575
|
+
{
|
|
1576
|
+
"backend": "maca",
|
|
1577
|
+
"backend_version": "3.3",
|
|
1578
|
+
"original_backend_version": "3.3.0",
|
|
1579
|
+
"backend_variant": "",
|
|
1580
|
+
"service": "sglang",
|
|
1581
|
+
"service_version": "0.5.6",
|
|
1582
|
+
"platform": "linux/amd64",
|
|
1583
|
+
"docker_image": "gpustack/runner:maca3.3-sglang0.5.6",
|
|
1584
|
+
"deprecated": false
|
|
1585
|
+
},
|
|
1586
|
+
{
|
|
1587
|
+
"backend": "maca",
|
|
1588
|
+
"backend_version": "3.3",
|
|
1589
|
+
"original_backend_version": "3.3.0",
|
|
1590
|
+
"backend_variant": "",
|
|
1591
|
+
"service": "vllm",
|
|
1592
|
+
"service_version": "0.11.2",
|
|
1593
|
+
"platform": "linux/amd64",
|
|
1594
|
+
"docker_image": "gpustack/runner:maca3.3-vllm0.11.2",
|
|
1595
|
+
"deprecated": false
|
|
1596
|
+
},
|
|
1421
1597
|
{
|
|
1422
1598
|
"backend": "maca",
|
|
1423
1599
|
"backend_version": "3.2",
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.4
|
|
2
2
|
Name: gpustack-runner
|
|
3
|
-
Version: 0.1.24
|
|
3
|
+
Version: 0.1.24.post2
|
|
4
4
|
Summary: GPUStack Runner is library for registering runnable accelerated backends and services in GPUStack.
|
|
5
5
|
Project-URL: Homepage, https://github.com/gpustack/runner
|
|
6
6
|
Project-URL: Bug Tracker, https://github.com/gpustack/gpustack/issues
|
|
@@ -46,24 +46,23 @@ The following table lists the supported accelerated backends and their correspon
|
|
|
46
46
|
|
|
47
47
|
### Ascend CANN
|
|
48
48
|
|
|
49
|
-
> [!CAUTION]
|
|
50
|
-
> Since v0.1.23:
|
|
51
|
-
> - Deprecated MindIE `2.1.rc1`.
|
|
52
|
-
|
|
53
49
|
> [!WARNING]
|
|
54
50
|
> - The Atlas 300I series is currently experimental in vLLM, only supporting eager mode and float16 data type. And there
|
|
55
51
|
are some known issues for running vLLM, you can refer to
|
|
56
52
|
vllm-ascend [#3316](https://github.com/vllm-project/vllm-ascend/issues/3316)
|
|
57
53
|
and [#2795](https://github.com/vllm-project/vllm-ascend/issues/2795).
|
|
58
54
|
|
|
59
|
-
| CANN Version <br/> (Variant) | MindIE
|
|
60
|
-
|
|
61
|
-
| 8.
|
|
62
|
-
| 8.
|
|
63
|
-
| 8.
|
|
64
|
-
| 8.
|
|
65
|
-
| 8.
|
|
66
|
-
| 8.
|
|
55
|
+
| CANN Version <br/> (Variant) | MindIE | vLLM | SGLang |
|
|
56
|
+
|------------------------------|-----------|------------------------------------------------------------|------------------------|
|
|
57
|
+
| 8.5 (A3/910C) | `2.3.0` | `0.14.1`, `0.13.0` | `0.5.8` |
|
|
58
|
+
| 8.5 (910B) | `2.3.0` | `0.14.1`, `0.13.0` | `0.5.8` |
|
|
59
|
+
| 8.5 (310P) | `2.3.0` | `0.14.1` | |
|
|
60
|
+
| 8.3 (A3/910C) | `2.2.rc1` | `0.12.0`, `0.11.0` | `0.5.7`, `0.5.6.post2` |
|
|
61
|
+
| 8.3 (910B) | `2.2.rc1` | `0.12.0`, `0.11.0` | `0.5.7`, `0.5.6.post2` |
|
|
62
|
+
| 8.3 (310P) | `2.2.rc1` | | |
|
|
63
|
+
| 8.2 (A3/910C) | `2.1.rc2` | `0.10.2`, `0.10.1.1` | `0.5.2`, `0.5.1.post3` |
|
|
64
|
+
| 8.2 (910B) | `2.1.rc2` | `0.10.2`, `0.10.1.1`, <br/>`0.10.0`, `0.9.2`, <br/>`0.9.1` | `0.5.2`, `0.5.1.post3` |
|
|
65
|
+
| 8.2 (310P) | `2.1.rc2` | `0.10.0`, `0.9.2` | |
|
|
67
66
|
|
|
68
67
|
### Iluvatar CoreX
|
|
69
68
|
|
|
@@ -73,13 +72,6 @@ The following table lists the supported accelerated backends and their correspon
|
|
|
73
72
|
|
|
74
73
|
### NVIDIA CUDA
|
|
75
74
|
|
|
76
|
-
> [!CAUTION]
|
|
77
|
-
> Since v0.1.23:
|
|
78
|
-
> - Deprecated all services for CUDA 12.4.
|
|
79
|
-
> - Deprecated vLLM `0.11.0`, `0.10.1.1`, `0.10.0`.
|
|
80
|
-
> - Deprecated SGLang `0.5.5`.
|
|
81
|
-
> - Deprecated VoxBox `0.0.20`.
|
|
82
|
-
|
|
83
75
|
> [!NOTE]
|
|
84
76
|
> - CUDA 12.9 supports Compute Capabilities:
|
|
85
77
|
`7.5 8.0+PTX 8.9 9.0 10.0 10.3 12.0 12.1+PTX`.
|
|
@@ -88,12 +80,11 @@ The following table lists the supported accelerated backends and their correspon
|
|
|
88
80
|
> - CUDA 12.6/12.4 supports Compute Capabilities:
|
|
89
81
|
`7.5 8.0+PTX 8.9 9.0+PTX`.
|
|
90
82
|
|
|
91
|
-
| CUDA Version <br/> (Variant) | vLLM
|
|
92
|
-
|
|
93
|
-
| 12.9 | `0.13.0`, `0.12.0`, <br/>`0.11.2`
|
|
94
|
-
| 12.8 | `0.13.0`, `0.12.0`, <br/>`0.11.2`,
|
|
95
|
-
| 12.6 | `0.13.0`, `0.12.0`, <br/>`0.11.2
|
|
96
|
-
| 12.4 | ~~`0.11.0`~~, ~~`0.10.2`~~, <br/>~~`0.10.1.1`~~, ~~`0.10.0`~~ | | ~~`0.0.20`~~ |
|
|
83
|
+
| CUDA Version <br/> (Variant) | vLLM | SGLang | VoxBox |
|
|
84
|
+
|------------------------------|---------------------------------------------|-----------------------------------------------------------|----------|
|
|
85
|
+
| 12.9 | `0.13.0`, `0.12.0`, <br/>`0.11.2` | `0.5.7`, `0.5.6.post2` | |
|
|
86
|
+
| 12.8 | `0.13.0`, `0.12.0`, <br/>`0.11.2`, `0.10.2` | `0.5.7`, `0.5.6.post2`, <br/>`0.5.5.post3`, `0.5.4.post3` | `0.0.21` |
|
|
87
|
+
| 12.6 | `0.13.0`, `0.12.0`, <br/>`0.11.2`,`0.10.2` | | `0.0.21` |
|
|
97
88
|
|
|
98
89
|
### Hygon DTK
|
|
99
90
|
|
|
@@ -101,7 +92,7 @@ The following table lists the supported accelerated backends and their correspon
|
|
|
101
92
|
|-----------------------------|----------------------------|
|
|
102
93
|
| 25.04 | `0.11.0`, `0.9.2`, `0.8.5` |
|
|
103
94
|
|
|
104
|
-
###
|
|
95
|
+
### T-Head HGGC
|
|
105
96
|
|
|
106
97
|
| HGGC Version <br/> (Variant) | vLLM | SGLang |
|
|
107
98
|
|------------------------------|----------|---------|
|
|
@@ -109,10 +100,11 @@ The following table lists the supported accelerated backends and their correspon
|
|
|
109
100
|
|
|
110
101
|
### MetaX MACA
|
|
111
102
|
|
|
112
|
-
| MACA Version <br/> (Variant) | vLLM |
|
|
113
|
-
|
|
114
|
-
| 3.
|
|
115
|
-
| 3.
|
|
103
|
+
| MACA Version <br/> (Variant) | vLLM | SGLang |
|
|
104
|
+
|------------------------------|----------|---------|
|
|
105
|
+
| 3.3 | `0.11.2` | `0.5.6` |
|
|
106
|
+
| 3.2 | `0.10.2` | |
|
|
107
|
+
| 3.0 | `0.9.1` | |
|
|
116
108
|
|
|
117
109
|
### MThreads MUSA
|
|
118
110
|
|
|
@@ -123,29 +115,23 @@ The following table lists the supported accelerated backends and their correspon
|
|
|
123
115
|
|
|
124
116
|
### AMD ROCm
|
|
125
117
|
|
|
126
|
-
> [!CAUTION]
|
|
127
|
-
> Since v0.1.23:
|
|
128
|
-
> - Deprecated all services for ROCm 6.3.
|
|
129
|
-
> - Deprecated vLLM `0.11.0`.
|
|
130
|
-
|
|
131
118
|
> [!NOTE]
|
|
132
119
|
> - ROCm 7.0 supports LLVM targets:
|
|
133
120
|
`gfx908 gfx90a gfx942 gfx950 gfx1030 gfx1100 gfx1101 gfx1200 gfx1201 gfx1150 gfx1151`.
|
|
134
|
-
> - ROCm 6.4
|
|
121
|
+
> - ROCm 6.4 supports LLVM targets:
|
|
135
122
|
`gfx908 gfx90a gfx942 gfx1030 gfx1100`.
|
|
136
123
|
|
|
137
124
|
> [!WARNING]
|
|
138
|
-
> - ROCm 7.0 vLLM `0.11.2
|
|
139
|
-
7.0 specific PyTorch build. Although supports ROCm 7.0 in vLLM `0.11.2
|
|
125
|
+
> - ROCm 7.0 vLLM `0.11.2` are reusing the official ROCm 6.4 PyTorch 2.9 wheel package rather than a ROCm
|
|
126
|
+
7.0 specific PyTorch build. Although supports ROCm 7.0 in vLLM `0.11.2`, `gfx1150/gfx1151` are not supported yet.
|
|
140
127
|
> - ROCm 6.4 vLLM `0.13.0` supports `gfx903 gfx90a gfx942` only.
|
|
141
128
|
> - ROCm 6.4 SGLang supports `gfx942` only.
|
|
142
129
|
> - ROCm 7.0 SGLang supports `gfx950` only.
|
|
143
130
|
|
|
144
|
-
| ROCm Version <br/> (Variant) | vLLM
|
|
145
|
-
|
|
146
|
-
| 7.0 | `0.13.0`, `0.12.0`, <br/>`0.11.2
|
|
147
|
-
| 6.4 | `0.13.0`, `0.12.0`, <br/>`0.11.2`, `0.10.2`
|
|
148
|
-
| 6.3 | ~~`0.10.1.1`~~, ~~`0.10.0`~~ | |
|
|
131
|
+
| ROCm Version <br/> (Variant) | vLLM | SGLang |
|
|
132
|
+
|------------------------------|---------------------------------------------|--------------------------------------------|
|
|
133
|
+
| 7.0 | `0.13.0`, `0.12.0`, <br/>`0.11.2` | `0.5.7`, `0.5.6.post2` |
|
|
134
|
+
| 6.4 | `0.13.0`, `0.12.0`, <br/>`0.11.2`, `0.10.2` | `0.5.7`, `0.5.6.post2`, <br/>`0.5.5.post3` |
|
|
149
135
|
|
|
150
136
|
## Directory Structure
|
|
151
137
|
|
|
@@ -0,0 +1,17 @@
|
|
|
1
|
+
gpustack_runner/__init__.py,sha256=ss8_wsk1oo13qFlP9usrXKX4ypHs1NZb2OVd6uFx5cc,715
|
|
2
|
+
gpustack_runner/__main__.py,sha256=uvpk9GtyJGtHaKM7DyM64N5mwehwpDD3v8ba6Yy8V3A,1364
|
|
3
|
+
gpustack_runner/__utils__.py,sha256=LSo0Iqxd5OjQFncVOYcqC8cncjtywf0qac31UPw7Ou4,4372
|
|
4
|
+
gpustack_runner/_version.py,sha256=SigvJTYuRQZ32Z9cgZmeBjPHmNv3OPkdCkWmY-EdHtc,792
|
|
5
|
+
gpustack_runner/_version.pyi,sha256=A42NoSgcqEXVy2OeNm4LXC9CbyonbooYrSUBlPm2lGY,156
|
|
6
|
+
gpustack_runner/envs.py,sha256=Wm0GTIiDJIT1zEjLpaPZNLbOs23NNFc2Y6zZuRLlTKQ,3470
|
|
7
|
+
gpustack_runner/runner.py,sha256=Rk4nyHj7Bn7ibAdxspXpXplMgWVMsG9Jb7GV8sHk2Ig,26310
|
|
8
|
+
gpustack_runner/runner.py.json,sha256=PJ0Tc0cxlvn0qOHUYqiEYvCBO5p9P_zKGetE7sOslOc,50303
|
|
9
|
+
gpustack_runner/cmds/__init__.py,sha256=zjdv_OC674KAcitjiHrHbXnAwLtw8Ju3psW0IKFqPIg,471
|
|
10
|
+
gpustack_runner/cmds/__types__.py,sha256=7C4kQM0EHPD8WpJpTo6kh9rEdkrYALcLQ-GAzMMsqV8,789
|
|
11
|
+
gpustack_runner/cmds/images.py,sha256=qkL862Xx6pcrLALhB8azMmtOAlIT-V5iGPHgSllnsPs,49029
|
|
12
|
+
gpustack_runner/_version_appendix.py,sha256=V3FyUmXPRZnROBj6MvdU2i0qBK2IazwbxkITbNVc7tQ,23
|
|
13
|
+
gpustack_runner-0.1.24.post2.dist-info/METADATA,sha256=-Zmb7jAVyM7KjyoFUesulwabed6RUnX57XVEfO7oA0A,12804
|
|
14
|
+
gpustack_runner-0.1.24.post2.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
|
|
15
|
+
gpustack_runner-0.1.24.post2.dist-info/entry_points.txt,sha256=M1Dxl6cY0kIgf2I4pPsV-_kU6BAtjj93spmsXAdwW3s,66
|
|
16
|
+
gpustack_runner-0.1.24.post2.dist-info/licenses/LICENSE,sha256=OiPibowBvB-NHV3TP_NOj18XNBlXcshXZFMpa3uvKVE,10362
|
|
17
|
+
gpustack_runner-0.1.24.post2.dist-info/RECORD,,
|
|
@@ -1,17 +0,0 @@
|
|
|
1
|
-
gpustack_runner/__init__.py,sha256=ss8_wsk1oo13qFlP9usrXKX4ypHs1NZb2OVd6uFx5cc,715
|
|
2
|
-
gpustack_runner/__main__.py,sha256=uvpk9GtyJGtHaKM7DyM64N5mwehwpDD3v8ba6Yy8V3A,1364
|
|
3
|
-
gpustack_runner/__utils__.py,sha256=LSo0Iqxd5OjQFncVOYcqC8cncjtywf0qac31UPw7Ou4,4372
|
|
4
|
-
gpustack_runner/_version.py,sha256=PGtcaWdpQHoovDMPesJn_siACNRb_TCcIlFvVB6YEig,777
|
|
5
|
-
gpustack_runner/_version.pyi,sha256=A42NoSgcqEXVy2OeNm4LXC9CbyonbooYrSUBlPm2lGY,156
|
|
6
|
-
gpustack_runner/envs.py,sha256=Wm0GTIiDJIT1zEjLpaPZNLbOs23NNFc2Y6zZuRLlTKQ,3470
|
|
7
|
-
gpustack_runner/runner.py,sha256=WOERxhW2huq_zESGaJ7ntlWFeQJkOs1NWc8XSrlTJTI,26594
|
|
8
|
-
gpustack_runner/runner.py.json,sha256=rQEZrBRAQngUY9z2AntWTnnwjhTzZI_yywMxhjAAcAw,45403
|
|
9
|
-
gpustack_runner/cmds/__init__.py,sha256=zjdv_OC674KAcitjiHrHbXnAwLtw8Ju3psW0IKFqPIg,471
|
|
10
|
-
gpustack_runner/cmds/__types__.py,sha256=7C4kQM0EHPD8WpJpTo6kh9rEdkrYALcLQ-GAzMMsqV8,789
|
|
11
|
-
gpustack_runner/cmds/images.py,sha256=Bd_yoW7MboVNAxtpMlf8JKm9vPGvuIcHcxInA9VCUsI,49223
|
|
12
|
-
gpustack_runner/_version_appendix.py,sha256=ZgtpyK-U9XcDUsfZR8KHFW6SDKmtMp9w4_lXLNiK5YY,23
|
|
13
|
-
gpustack_runner-0.1.24.dist-info/METADATA,sha256=T6nsHWr00ZtYpB8Ri0ADF5aqhTJxc2V54QKzi_B71mw,13559
|
|
14
|
-
gpustack_runner-0.1.24.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
|
|
15
|
-
gpustack_runner-0.1.24.dist-info/entry_points.txt,sha256=M1Dxl6cY0kIgf2I4pPsV-_kU6BAtjj93spmsXAdwW3s,66
|
|
16
|
-
gpustack_runner-0.1.24.dist-info/licenses/LICENSE,sha256=OiPibowBvB-NHV3TP_NOj18XNBlXcshXZFMpa3uvKVE,10362
|
|
17
|
-
gpustack_runner-0.1.24.dist-info/RECORD,,
|
|
File without changes
|
{gpustack_runner-0.1.24.dist-info → gpustack_runner-0.1.24.post2.dist-info}/entry_points.txt
RENAMED
|
File without changes
|
{gpustack_runner-0.1.24.dist-info → gpustack_runner-0.1.24.post2.dist-info}/licenses/LICENSE
RENAMED
|
File without changes
|