xinference 0.15.3__py3-none-any.whl → 0.15.4__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.
Potentially problematic release.
This version of xinference might be problematic. Click here for more details.
- xinference/_version.py +3 -3
- xinference/api/restful_api.py +29 -2
- xinference/client/restful/restful_client.py +10 -0
- xinference/constants.py +4 -0
- xinference/core/image_interface.py +76 -23
- xinference/core/model.py +80 -39
- xinference/core/progress_tracker.py +187 -0
- xinference/core/supervisor.py +11 -0
- xinference/core/worker.py +1 -0
- xinference/model/audio/chattts.py +2 -1
- xinference/model/audio/core.py +0 -2
- xinference/model/audio/model_spec.json +8 -0
- xinference/model/audio/model_spec_modelscope.json +9 -0
- xinference/model/image/core.py +6 -7
- xinference/model/image/sdapi.py +35 -4
- xinference/model/image/stable_diffusion/core.py +208 -78
- xinference/model/llm/llm_family.json +16 -16
- xinference/model/llm/llm_family_modelscope.json +16 -12
- xinference/model/llm/transformers/cogvlm2.py +2 -1
- xinference/model/llm/transformers/cogvlm2_video.py +2 -0
- xinference/model/llm/transformers/core.py +6 -2
- xinference/model/llm/transformers/deepseek_vl.py +2 -0
- xinference/model/llm/transformers/glm4v.py +2 -1
- xinference/model/llm/transformers/intern_vl.py +2 -0
- xinference/model/llm/transformers/minicpmv25.py +2 -0
- xinference/model/llm/transformers/minicpmv26.py +2 -0
- xinference/model/llm/transformers/omnilmm.py +2 -0
- xinference/model/llm/transformers/qwen2_audio.py +11 -4
- xinference/model/llm/transformers/qwen2_vl.py +2 -28
- xinference/model/llm/transformers/qwen_vl.py +2 -1
- xinference/model/llm/transformers/utils.py +35 -2
- xinference/model/llm/transformers/yi_vl.py +2 -0
- xinference/model/llm/utils.py +58 -14
- xinference/model/llm/vllm/core.py +52 -8
- xinference/model/llm/vllm/utils.py +0 -1
- xinference/model/utils.py +7 -4
- xinference/model/video/core.py +0 -2
- {xinference-0.15.3.dist-info → xinference-0.15.4.dist-info}/METADATA +3 -3
- {xinference-0.15.3.dist-info → xinference-0.15.4.dist-info}/RECORD +43 -42
- {xinference-0.15.3.dist-info → xinference-0.15.4.dist-info}/LICENSE +0 -0
- {xinference-0.15.3.dist-info → xinference-0.15.4.dist-info}/WHEEL +0 -0
- {xinference-0.15.3.dist-info → xinference-0.15.4.dist-info}/entry_points.txt +0 -0
- {xinference-0.15.3.dist-info → xinference-0.15.4.dist-info}/top_level.txt +0 -0
xinference/model/llm/utils.py
CHANGED
|
@@ -29,6 +29,7 @@ from ...types import (
|
|
|
29
29
|
ChatCompletion,
|
|
30
30
|
ChatCompletionChoice,
|
|
31
31
|
ChatCompletionChunk,
|
|
32
|
+
ChatCompletionMessage,
|
|
32
33
|
Completion,
|
|
33
34
|
CompletionChoice,
|
|
34
35
|
CompletionChunk,
|
|
@@ -50,6 +51,7 @@ QWEN_TOOL_CALL_FAMILY = [
|
|
|
50
51
|
"qwen1.5-moe-chat",
|
|
51
52
|
"qwen2-instruct",
|
|
52
53
|
"qwen2-moe-instruct",
|
|
54
|
+
"qwen2.5-instruct",
|
|
53
55
|
]
|
|
54
56
|
|
|
55
57
|
GLM4_TOOL_CALL_FAMILY = [
|
|
@@ -57,6 +59,10 @@ GLM4_TOOL_CALL_FAMILY = [
|
|
|
57
59
|
"glm4-chat-1m",
|
|
58
60
|
]
|
|
59
61
|
|
|
62
|
+
LLAMA3_TOOL_CALL_FAMILY = [
|
|
63
|
+
"llama-3.1-instruct",
|
|
64
|
+
]
|
|
65
|
+
|
|
60
66
|
QWEN_TOOL_CALL_SYMBOLS = ["<tool_call>", "</tool_call>"]
|
|
61
67
|
|
|
62
68
|
|
|
@@ -113,7 +119,7 @@ class ChatModelMixin:
|
|
|
113
119
|
return self._build_from_raw_template(messages, chat_template, **kwargs)
|
|
114
120
|
|
|
115
121
|
@staticmethod
|
|
116
|
-
def get_specific_prompt(model_family: str, messages: List[
|
|
122
|
+
def get_specific_prompt(model_family: str, messages: List[ChatCompletionMessage]):
|
|
117
123
|
"""
|
|
118
124
|
Inspired by FastChat. Format chat history into a prompt according to the prompty style of
|
|
119
125
|
different models.
|
|
@@ -129,7 +135,7 @@ class ChatModelMixin:
|
|
|
129
135
|
ret = (
|
|
130
136
|
"<s>"
|
|
131
137
|
if system_prompt == ""
|
|
132
|
-
else "<s><|im_start|>system\n"
|
|
138
|
+
else "<s><|im_start|>system\n" # type: ignore
|
|
133
139
|
+ system_prompt
|
|
134
140
|
+ intra_message_sep
|
|
135
141
|
+ "\n"
|
|
@@ -333,8 +339,9 @@ class ChatModelMixin:
|
|
|
333
339
|
for content in contents:
|
|
334
340
|
content = content.strip()
|
|
335
341
|
if content:
|
|
336
|
-
|
|
337
|
-
|
|
342
|
+
pos = content.find(QWEN_TOOL_CALL_SYMBOLS[0])
|
|
343
|
+
if pos != -1:
|
|
344
|
+
content = content[pos + len(QWEN_TOOL_CALL_SYMBOLS[0]) :]
|
|
338
345
|
content = content.strip()
|
|
339
346
|
try:
|
|
340
347
|
res = json.loads(content)
|
|
@@ -353,6 +360,15 @@ class ChatModelMixin:
|
|
|
353
360
|
text = c["choices"][0]["text"]
|
|
354
361
|
return cls._handle_qwen_tool_result(text)
|
|
355
362
|
|
|
363
|
+
@classmethod
|
|
364
|
+
def _eval_llama3_chat_arguments(cls, c) -> List[Tuple]:
|
|
365
|
+
text = c["choices"][0]["text"]
|
|
366
|
+
try:
|
|
367
|
+
data = eval(text, {}, {})
|
|
368
|
+
return [(None, data["name"], data["parameters"])]
|
|
369
|
+
except Exception:
|
|
370
|
+
return [(text, None, None)]
|
|
371
|
+
|
|
356
372
|
@classmethod
|
|
357
373
|
def _eval_tool_arguments(cls, model_family, c):
|
|
358
374
|
family = model_family.model_family or model_family.model_name
|
|
@@ -360,6 +376,8 @@ class ChatModelMixin:
|
|
|
360
376
|
result = cls._eval_glm_chat_arguments(c)
|
|
361
377
|
elif family in QWEN_TOOL_CALL_FAMILY:
|
|
362
378
|
result = cls._eval_qwen_chat_arguments(c)
|
|
379
|
+
elif family in LLAMA3_TOOL_CALL_FAMILY:
|
|
380
|
+
result = cls._eval_llama3_chat_arguments(c)
|
|
363
381
|
else:
|
|
364
382
|
raise Exception(
|
|
365
383
|
f"Model {model_family.model_name} is not support tool calls."
|
|
@@ -376,16 +394,14 @@ class ChatModelMixin:
|
|
|
376
394
|
for content, func, args in tool_result:
|
|
377
395
|
if func:
|
|
378
396
|
tool_calls.append(
|
|
379
|
-
|
|
380
|
-
{
|
|
381
|
-
|
|
382
|
-
|
|
383
|
-
"
|
|
384
|
-
|
|
385
|
-
|
|
386
|
-
|
|
387
|
-
}
|
|
388
|
-
]
|
|
397
|
+
{
|
|
398
|
+
"id": f"call_{_id}",
|
|
399
|
+
"type": "function",
|
|
400
|
+
"function": {
|
|
401
|
+
"name": func,
|
|
402
|
+
"arguments": json.dumps(args, ensure_ascii=False),
|
|
403
|
+
},
|
|
404
|
+
}
|
|
389
405
|
)
|
|
390
406
|
else:
|
|
391
407
|
failed_contents.append(content)
|
|
@@ -471,6 +487,34 @@ class ChatModelMixin:
|
|
|
471
487
|
"usage": usage,
|
|
472
488
|
}
|
|
473
489
|
|
|
490
|
+
def _transform_messages(
|
|
491
|
+
self,
|
|
492
|
+
messages: List[ChatCompletionMessage],
|
|
493
|
+
):
|
|
494
|
+
transformed_messages = []
|
|
495
|
+
for msg in messages:
|
|
496
|
+
new_content = []
|
|
497
|
+
role = msg["role"]
|
|
498
|
+
content = msg["content"]
|
|
499
|
+
if isinstance(content, str):
|
|
500
|
+
new_content.append({"type": "text", "text": content})
|
|
501
|
+
elif isinstance(content, List):
|
|
502
|
+
for item in content: # type: ignore
|
|
503
|
+
if "text" in item:
|
|
504
|
+
new_content.append({"type": "text", "text": item["text"]})
|
|
505
|
+
elif "image_url" in item:
|
|
506
|
+
new_content.append(
|
|
507
|
+
{"type": "image", "image": item["image_url"]["url"]}
|
|
508
|
+
)
|
|
509
|
+
elif "video_url" in item:
|
|
510
|
+
new_content.append(
|
|
511
|
+
{"type": "video", "video": item["video_url"]["url"]}
|
|
512
|
+
)
|
|
513
|
+
new_message = {"role": role, "content": new_content}
|
|
514
|
+
transformed_messages.append(new_message)
|
|
515
|
+
|
|
516
|
+
return transformed_messages
|
|
517
|
+
|
|
474
518
|
|
|
475
519
|
def get_file_location(
|
|
476
520
|
llm_family: LLMFamilyV1, spec: LLMSpecV1, quantization: str
|
|
@@ -34,6 +34,7 @@ from typing import (
|
|
|
34
34
|
from ....types import (
|
|
35
35
|
ChatCompletion,
|
|
36
36
|
ChatCompletionChunk,
|
|
37
|
+
ChatCompletionMessage,
|
|
37
38
|
Completion,
|
|
38
39
|
CompletionChoice,
|
|
39
40
|
CompletionChunk,
|
|
@@ -175,6 +176,9 @@ if VLLM_INSTALLED and vllm.__version__ > "0.5.3":
|
|
|
175
176
|
if VLLM_INSTALLED and vllm.__version__ >= "0.6.1":
|
|
176
177
|
VLLM_SUPPORTED_VISION_MODEL_LIST.append("internvl2")
|
|
177
178
|
|
|
179
|
+
if VLLM_INSTALLED and vllm.__version__ >= "0.6.3":
|
|
180
|
+
VLLM_SUPPORTED_VISION_MODEL_LIST.append("qwen2-vl-instruct")
|
|
181
|
+
|
|
178
182
|
|
|
179
183
|
class VLLMModel(LLM):
|
|
180
184
|
def __init__(
|
|
@@ -309,11 +313,6 @@ class VLLMModel(LLM):
|
|
|
309
313
|
model_config.setdefault("max_num_seqs", 256)
|
|
310
314
|
model_config.setdefault("quantization", None)
|
|
311
315
|
model_config.setdefault("max_model_len", None)
|
|
312
|
-
model_config["limit_mm_per_prompt"] = (
|
|
313
|
-
json.loads(model_config.get("limit_mm_per_prompt")) # type: ignore
|
|
314
|
-
if model_config.get("limit_mm_per_prompt")
|
|
315
|
-
else None
|
|
316
|
-
)
|
|
317
316
|
|
|
318
317
|
return model_config
|
|
319
318
|
|
|
@@ -733,6 +732,33 @@ class VLLMVisionModel(VLLMModel, ChatModelMixin):
|
|
|
733
732
|
return False
|
|
734
733
|
return VLLM_INSTALLED
|
|
735
734
|
|
|
735
|
+
def _sanitize_model_config(
|
|
736
|
+
self, model_config: Optional[VLLMModelConfig]
|
|
737
|
+
) -> VLLMModelConfig:
|
|
738
|
+
if model_config is None:
|
|
739
|
+
model_config = VLLMModelConfig()
|
|
740
|
+
|
|
741
|
+
cuda_count = self._get_cuda_count()
|
|
742
|
+
|
|
743
|
+
model_config.setdefault("tokenizer_mode", "auto")
|
|
744
|
+
model_config.setdefault("trust_remote_code", True)
|
|
745
|
+
model_config.setdefault("tensor_parallel_size", cuda_count)
|
|
746
|
+
model_config.setdefault("block_size", 16)
|
|
747
|
+
model_config.setdefault("swap_space", 4)
|
|
748
|
+
model_config.setdefault("gpu_memory_utilization", 0.90)
|
|
749
|
+
model_config.setdefault("max_num_seqs", 256)
|
|
750
|
+
model_config.setdefault("quantization", None)
|
|
751
|
+
model_config.setdefault("max_model_len", None)
|
|
752
|
+
model_config["limit_mm_per_prompt"] = (
|
|
753
|
+
json.loads(model_config.get("limit_mm_per_prompt")) # type: ignore
|
|
754
|
+
if model_config.get("limit_mm_per_prompt")
|
|
755
|
+
else {
|
|
756
|
+
"image": 2, # default 2 images all chat
|
|
757
|
+
}
|
|
758
|
+
)
|
|
759
|
+
|
|
760
|
+
return model_config
|
|
761
|
+
|
|
736
762
|
def _sanitize_chat_config(
|
|
737
763
|
self,
|
|
738
764
|
generate_config: Optional[Dict] = None,
|
|
@@ -755,14 +781,32 @@ class VLLMVisionModel(VLLMModel, ChatModelMixin):
|
|
|
755
781
|
@vllm_check
|
|
756
782
|
async def async_chat(
|
|
757
783
|
self,
|
|
758
|
-
messages: List[
|
|
784
|
+
messages: List[ChatCompletionMessage], # type: ignore
|
|
759
785
|
generate_config: Optional[Dict] = None,
|
|
760
786
|
request_id: Optional[str] = None,
|
|
761
787
|
) -> Union[ChatCompletion, AsyncGenerator[ChatCompletionChunk, None]]:
|
|
788
|
+
messages = self._transform_messages(messages)
|
|
789
|
+
tools = generate_config.pop("tools", []) if generate_config else None
|
|
790
|
+
|
|
762
791
|
model_family = self.model_family.model_family or self.model_family.model_name
|
|
763
|
-
prompt, images = self.get_specific_prompt(model_family, messages)
|
|
764
792
|
|
|
765
|
-
if
|
|
793
|
+
if "internvl2" not in model_family.lower():
|
|
794
|
+
from qwen_vl_utils import process_vision_info
|
|
795
|
+
|
|
796
|
+
full_context_kwargs = {}
|
|
797
|
+
if tools and model_family in QWEN_TOOL_CALL_FAMILY:
|
|
798
|
+
full_context_kwargs["tools"] = tools
|
|
799
|
+
assert self.model_family.chat_template is not None
|
|
800
|
+
prompt = self.get_full_context(
|
|
801
|
+
messages, self.model_family.chat_template, **full_context_kwargs
|
|
802
|
+
)
|
|
803
|
+
images, video_inputs = process_vision_info(messages)
|
|
804
|
+
if video_inputs:
|
|
805
|
+
raise ValueError("Not support video input now.")
|
|
806
|
+
else:
|
|
807
|
+
prompt, images = self.get_specific_prompt(model_family, messages)
|
|
808
|
+
|
|
809
|
+
if not images:
|
|
766
810
|
inputs = {
|
|
767
811
|
"prompt": prompt,
|
|
768
812
|
}
|
xinference/model/utils.py
CHANGED
|
@@ -23,12 +23,15 @@ import huggingface_hub
|
|
|
23
23
|
import numpy as np
|
|
24
24
|
import torch
|
|
25
25
|
|
|
26
|
-
from ..constants import
|
|
26
|
+
from ..constants import (
|
|
27
|
+
XINFERENCE_CACHE_DIR,
|
|
28
|
+
XINFERENCE_DOWNLOAD_MAX_ATTEMPTS,
|
|
29
|
+
XINFERENCE_ENV_MODEL_SRC,
|
|
30
|
+
)
|
|
27
31
|
from ..device_utils import get_available_device, is_device_available
|
|
28
32
|
from .core import CacheableModelSpec
|
|
29
33
|
|
|
30
34
|
logger = logging.getLogger(__name__)
|
|
31
|
-
MAX_ATTEMPTS = 3
|
|
32
35
|
IS_NEW_HUGGINGFACE_HUB: bool = huggingface_hub.__version__ >= "0.23.0"
|
|
33
36
|
|
|
34
37
|
|
|
@@ -100,11 +103,11 @@ def retry_download(
|
|
|
100
103
|
**kwargs,
|
|
101
104
|
):
|
|
102
105
|
last_ex = None
|
|
103
|
-
for current_attempt in range(1,
|
|
106
|
+
for current_attempt in range(1, XINFERENCE_DOWNLOAD_MAX_ATTEMPTS + 1):
|
|
104
107
|
try:
|
|
105
108
|
return download_func(*args, **kwargs)
|
|
106
109
|
except Exception as e:
|
|
107
|
-
remaining_attempts =
|
|
110
|
+
remaining_attempts = XINFERENCE_DOWNLOAD_MAX_ATTEMPTS - current_attempt
|
|
108
111
|
last_ex = e
|
|
109
112
|
logger.debug(
|
|
110
113
|
"Download failed: %s, download func: %s, download args: %s, kwargs: %s",
|
xinference/model/video/core.py
CHANGED
|
@@ -21,8 +21,6 @@ from ..core import CacheableModelSpec, ModelDescription
|
|
|
21
21
|
from ..utils import valid_model_revision
|
|
22
22
|
from .diffusers import DiffUsersVideoModel
|
|
23
23
|
|
|
24
|
-
MAX_ATTEMPTS = 3
|
|
25
|
-
|
|
26
24
|
logger = logging.getLogger(__name__)
|
|
27
25
|
|
|
28
26
|
MODEL_NAME_TO_REVISION: Dict[str, List[str]] = defaultdict(list)
|
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
Metadata-Version: 2.1
|
|
2
2
|
Name: xinference
|
|
3
|
-
Version: 0.15.
|
|
3
|
+
Version: 0.15.4
|
|
4
4
|
Summary: Model Serving Made Easy
|
|
5
5
|
Home-page: https://github.com/xorbitsai/inference
|
|
6
6
|
Author: Qin Xuye
|
|
@@ -71,7 +71,7 @@ Requires-Dist: WeTextProcessing<1.0.4; extra == "all"
|
|
|
71
71
|
Requires-Dist: librosa; extra == "all"
|
|
72
72
|
Requires-Dist: xxhash; extra == "all"
|
|
73
73
|
Requires-Dist: torchaudio; extra == "all"
|
|
74
|
-
Requires-Dist: ChatTTS
|
|
74
|
+
Requires-Dist: ChatTTS<0.2,>0.1; extra == "all"
|
|
75
75
|
Requires-Dist: lightning>=2.0.0; extra == "all"
|
|
76
76
|
Requires-Dist: hydra-core>=1.3.2; extra == "all"
|
|
77
77
|
Requires-Dist: inflect; extra == "all"
|
|
@@ -104,7 +104,7 @@ Requires-Dist: WeTextProcessing<1.0.4; extra == "audio"
|
|
|
104
104
|
Requires-Dist: librosa; extra == "audio"
|
|
105
105
|
Requires-Dist: xxhash; extra == "audio"
|
|
106
106
|
Requires-Dist: torchaudio; extra == "audio"
|
|
107
|
-
Requires-Dist: ChatTTS
|
|
107
|
+
Requires-Dist: ChatTTS<0.2,>0.1; extra == "audio"
|
|
108
108
|
Requires-Dist: tiktoken; extra == "audio"
|
|
109
109
|
Requires-Dist: torch>=2.0.0; extra == "audio"
|
|
110
110
|
Requires-Dist: lightning>=2.0.0; extra == "audio"
|
|
@@ -1,15 +1,15 @@
|
|
|
1
1
|
xinference/__init__.py,sha256=muQ9V9y11BcIqlZhhc06oDf193H7bwDIa8e_wSoDKI8,986
|
|
2
2
|
xinference/_compat.py,sha256=xFztCfyrq3O_4bssL_ygghYkfxicv_ZhiX2YDDWHf-k,3571
|
|
3
|
-
xinference/_version.py,sha256=
|
|
3
|
+
xinference/_version.py,sha256=rsb6h82zrpecXkwoQPZlv48UaKLjrNi7qg2qXrXoSZE,498
|
|
4
4
|
xinference/conftest.py,sha256=56HYQjsAJcQrpZSmskniPqH9dLoW-i3Oud6NVTtc4io,9752
|
|
5
|
-
xinference/constants.py,sha256=
|
|
5
|
+
xinference/constants.py,sha256=QsYxf86vqJVbn5L2SXH5hdyCdiZy7hRD-Qx5Sse4DqE,3758
|
|
6
6
|
xinference/device_utils.py,sha256=zswJiws3VyTIaNO8z-MOcsJH_UiPoePPiKK5zoNrjTA,3285
|
|
7
7
|
xinference/fields.py,sha256=0UtBFaDNzn1n9MRjyTkNrolsIML-TpZfudWOejqjni8,5245
|
|
8
8
|
xinference/isolation.py,sha256=uhkzVyL3fSYZSuFexkG6Jm-tRTC5I607uNg000BXAnE,1949
|
|
9
9
|
xinference/types.py,sha256=LHTbNLf0zI-FLruxRuBt2KMpk2P4eKpYdFvh2qzNTGI,12458
|
|
10
10
|
xinference/utils.py,sha256=VSOJMFd9H7kce98OtJZbcDjjpfzRpHAFs8WU0xXPBM8,717
|
|
11
11
|
xinference/api/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQAL1ms,581
|
|
12
|
-
xinference/api/restful_api.py,sha256=
|
|
12
|
+
xinference/api/restful_api.py,sha256=Gp_1fGYLximhr9yTqxvBv9O84HO47-tnTwA5h7o8Ff4,84506
|
|
13
13
|
xinference/api/oauth2/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQAL1ms,581
|
|
14
14
|
xinference/api/oauth2/auth_service.py,sha256=74JzB42fbbmBu4Q1dW3A9Fp_N7167KgRGB42Z0NHjAM,6119
|
|
15
15
|
xinference/api/oauth2/types.py,sha256=K923sv_XySIUtM2Eozl9IG082IJcDOS5SFLrPZ5ELBg,996
|
|
@@ -18,20 +18,21 @@ xinference/client/__init__.py,sha256=Gc4HOzAy_1cic5kXlso7hahYgw89CKvZSJDicEU461k
|
|
|
18
18
|
xinference/client/common.py,sha256=iciZRs5YjM2gYsXnwACPMaiBZp4_XpawWwfym0Iyu40,1617
|
|
19
19
|
xinference/client/handlers.py,sha256=OKl_i5FA341wsQf_0onSOPbbW6V861WJrSP7ghtDc8c,527
|
|
20
20
|
xinference/client/restful/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQAL1ms,581
|
|
21
|
-
xinference/client/restful/restful_client.py,sha256=
|
|
21
|
+
xinference/client/restful/restful_client.py,sha256=eTZf9M0GG6ZaShWhpY7O-yG0BH3ceKZZ-d-DuVDg55g,51189
|
|
22
22
|
xinference/core/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQAL1ms,581
|
|
23
23
|
xinference/core/cache_tracker.py,sha256=3ubjYCU5aZToSp2GEuzedECVrg-PR4kThTefrFUkb9g,6971
|
|
24
24
|
xinference/core/chat_interface.py,sha256=tM4hQPZ0UVcmE4_-auXWkq2z0rWmZBwwXNwbbl5zvGQ,20666
|
|
25
25
|
xinference/core/event.py,sha256=42F38H2WOl6aPxp2oxX6WNxHRRxbnvYRmbt4Ar7NP4U,1640
|
|
26
|
-
xinference/core/image_interface.py,sha256
|
|
26
|
+
xinference/core/image_interface.py,sha256=WsJjrcJG3itQJb-qiuZKWtK4_XLPPnxkEy9VCyZcQmw,13636
|
|
27
27
|
xinference/core/metrics.py,sha256=ScmTG15Uq3h_ob72ybZSMWdnk8P4sUZFcm60f4ikSXc,2631
|
|
28
|
-
xinference/core/model.py,sha256=
|
|
28
|
+
xinference/core/model.py,sha256=bwmiqRctnXbJlsHhS3O3JA5G0xRNNRd_HqlAGyRrzVo,32086
|
|
29
|
+
xinference/core/progress_tracker.py,sha256=LIF6CLIlnEoSBkuDCraJktDOzZ31mQ4HOo6EVr3KpQM,6453
|
|
29
30
|
xinference/core/resource.py,sha256=FQ0aRt3T4ZQo0P6CZZf5QUKHiCsr5llBvKb1f7wfnxg,1611
|
|
30
31
|
xinference/core/scheduler.py,sha256=qONNFqAlnYDcmmzPO5jfU-r0aZ1Lhhpn1oSaA5CAGTE,15485
|
|
31
32
|
xinference/core/status_guard.py,sha256=4an1KjUOhCStgRQUw1VSzXcycXUtvhxwiMREKKcl1UI,2828
|
|
32
|
-
xinference/core/supervisor.py,sha256=
|
|
33
|
+
xinference/core/supervisor.py,sha256=Wkjhk1tfRuhyQmcVNrHZApWO09MDA5-Uu4u2p1GBj3I,52964
|
|
33
34
|
xinference/core/utils.py,sha256=p3ptQMdzKu9WxdUJ2EdDTXvPDl53BGwiNuVWuhaE4EU,8536
|
|
34
|
-
xinference/core/worker.py,sha256=
|
|
35
|
+
xinference/core/worker.py,sha256=QhxVhpeKl-QYKA_77kUXTj5-rhodHAXlOhgtvqZiiRI,46329
|
|
35
36
|
xinference/deploy/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQAL1ms,581
|
|
36
37
|
xinference/deploy/cmdline.py,sha256=YNXbPIT9zJIp5EQzl_rH5KwDDYqBd2CbaOVF8hA0lws,48120
|
|
37
38
|
xinference/deploy/local.py,sha256=gcH6WfTxfhjvNkxxKZH3tcGtXV48BEPoaLWYztZHaeo,3954
|
|
@@ -42,16 +43,16 @@ xinference/deploy/test/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQ
|
|
|
42
43
|
xinference/deploy/test/test_cmdline.py,sha256=m8xDzjtDuAJy0QkvYVJIZDuTB29cmYBV0d231JyRCPU,7714
|
|
43
44
|
xinference/model/__init__.py,sha256=J7cdxnDxbr-4c8JREXWEIZ8MkB_VokXVaEOsI7ycjho,1101
|
|
44
45
|
xinference/model/core.py,sha256=WQakLJgxrJMbTGn9AVaw-Tas7QL5M8cJsuYpGgH-io8,4706
|
|
45
|
-
xinference/model/utils.py,sha256
|
|
46
|
+
xinference/model/utils.py,sha256=-axJ9I1IZ0li5Y2qzfFrsz8nMOGnoXXhmULu_e4aIKc,11196
|
|
46
47
|
xinference/model/audio/__init__.py,sha256=G4n-MyzdarFVOndPRkEyZZrCwqFIG8yIsky6_5dife0,3433
|
|
47
|
-
xinference/model/audio/chattts.py,sha256=
|
|
48
|
-
xinference/model/audio/core.py,sha256=
|
|
48
|
+
xinference/model/audio/chattts.py,sha256=EXAfNATwblcilMjU3ff1dzTDJkjwQMXBo6zZ_517Jvo,4659
|
|
49
|
+
xinference/model/audio/core.py,sha256=8rIyw0PLW2Py2-V7xsij4uQGGY39D_uq6uICuICobO8,6494
|
|
49
50
|
xinference/model/audio/cosyvoice.py,sha256=Enur1Y4Xa-mpr7wwnoXWwhyh7PUAjrHZ8DV91tTrpjE,6426
|
|
50
51
|
xinference/model/audio/custom.py,sha256=8GXBRmTtR-GY03-E91nlRGTIuabCRzlt20ecU6Un6Y8,4985
|
|
51
52
|
xinference/model/audio/fish_speech.py,sha256=v2WVEV-BLWnbiDvqrx8WTGE_YNKmd9QoAF1LZBXWxn0,7310
|
|
52
53
|
xinference/model/audio/funasr.py,sha256=65z7U7_F14CCP-jg6BpeY3_49FK7Y5OCRSzrhhsklCg,4075
|
|
53
|
-
xinference/model/audio/model_spec.json,sha256=
|
|
54
|
-
xinference/model/audio/model_spec_modelscope.json,sha256=
|
|
54
|
+
xinference/model/audio/model_spec.json,sha256=rBfDYgiZNI0d1t01emx_UosEqap4JxD1OUJoocmlEMI,5120
|
|
55
|
+
xinference/model/audio/model_spec_modelscope.json,sha256=Ul7_zy49N5zvio8-1WZEBFzZXS5_ueT49frWepMF8KY,2031
|
|
55
56
|
xinference/model/audio/utils.py,sha256=pwo5cHh8nvhyBa9f-17QaVpXMSjmbpGbPYKwBBtEhGM,717
|
|
56
57
|
xinference/model/audio/whisper.py,sha256=PQL7rebGC7WlIOItuDtjdEtSJtlhxFkolot-Fj-8uDU,7982
|
|
57
58
|
xinference/model/embedding/__init__.py,sha256=1GmvQsbeeVUT-VRaRGetf8UT4RQgLWIzfp5kfX5jw-k,3567
|
|
@@ -67,22 +68,22 @@ xinference/model/flexible/launchers/__init__.py,sha256=X8w_2hKuQ9H3f90XYK7H_AQU4
|
|
|
67
68
|
xinference/model/flexible/launchers/image_process_launcher.py,sha256=APbbHls0N9DpLFL6_qTexuc5o6bQAvdgJEAZWU4clyw,2510
|
|
68
69
|
xinference/model/flexible/launchers/transformers_launcher.py,sha256=OZeeogDfopRUGhulP4PRJ4fZEJ2D9cfv7lcC2qJBoDE,2012
|
|
69
70
|
xinference/model/image/__init__.py,sha256=80HBIbKh6lh-BgNaTo6k0TxxKjdG30bwHAdCiwVk6wk,3198
|
|
70
|
-
xinference/model/image/core.py,sha256=
|
|
71
|
+
xinference/model/image/core.py,sha256=qdqFMpPa2OSi0d5a4_iASEjL8s5vxxl1IRTJjmxfwO0,8959
|
|
71
72
|
xinference/model/image/custom.py,sha256=5gjujQpJVTJ-pVB5LzBo4-bWKKOHzFlRaoRKJ_CuIUg,3769
|
|
72
73
|
xinference/model/image/model_spec.json,sha256=JyXU-v4ysRT4yqwkmXgISY3uVWjeSiBLyH8fS7XO1_g,5368
|
|
73
74
|
xinference/model/image/model_spec_modelscope.json,sha256=r3_m9XZo1QZgmASg5navOPs0ivlft5wVPF1SpbAVNBg,4266
|
|
74
|
-
xinference/model/image/sdapi.py,sha256=
|
|
75
|
+
xinference/model/image/sdapi.py,sha256=Xgdtnvw4Xwj1Nc0cBoDo_ogH6E2mFJqLvX0jSxxgdnA,5936
|
|
75
76
|
xinference/model/image/utils.py,sha256=gxg8jJ2nYaDknzCcSC53WCy1slbB5aWU14AbJbfm6Z4,906
|
|
76
77
|
xinference/model/image/stable_diffusion/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQAL1ms,581
|
|
77
|
-
xinference/model/image/stable_diffusion/core.py,sha256=
|
|
78
|
+
xinference/model/image/stable_diffusion/core.py,sha256=NeCfkKjRdebgOOGMmg8DGVPR29D9vgX3V72qETgwPp0,24230
|
|
78
79
|
xinference/model/llm/__init__.py,sha256=elINGzzDXmSissG32UMG7BfMqNgXwSM7USBaTorGWSA,12428
|
|
79
80
|
xinference/model/llm/core.py,sha256=f4nKVPTAseivij6mbL2yXEKxrzllKm-_i2ttSqckTCg,8157
|
|
80
|
-
xinference/model/llm/llm_family.json,sha256=
|
|
81
|
+
xinference/model/llm/llm_family.json,sha256=9peVrsgESrC-HPsIXS7wDcTsz_oNcGwrsw-Jh_yDSLU,285589
|
|
81
82
|
xinference/model/llm/llm_family.py,sha256=eqeaHwLeS2TDB_ATf_h6YkH6OiyyF_4cSF_bOq3pTws,37432
|
|
82
83
|
xinference/model/llm/llm_family_csghub.json,sha256=zMKWbihsxQNVB1u5iKJbZUkbOfQ4IPNq1KQ-8IDPQQA,8759
|
|
83
|
-
xinference/model/llm/llm_family_modelscope.json,sha256=
|
|
84
|
+
xinference/model/llm/llm_family_modelscope.json,sha256=IcnRScrHkM4b6-rrpto4hCZ541rkUpdkaUa8wgSw8No,215985
|
|
84
85
|
xinference/model/llm/memory.py,sha256=NEIMw6wWaF9S_bnBYq-EyuDhVbUEEeceQhwE1iwsrhI,10207
|
|
85
|
-
xinference/model/llm/utils.py,sha256=
|
|
86
|
+
xinference/model/llm/utils.py,sha256=_FHJHZ9d1tYj4NwiG4TYftEp9L5vah6slUkqHKnn21U,23543
|
|
86
87
|
xinference/model/llm/llama_cpp/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQAL1ms,581
|
|
87
88
|
xinference/model/llm/llama_cpp/core.py,sha256=vjuTapwbn-ZjUX-8WA0nFyicE4UGUSehU_csSetvcZw,10928
|
|
88
89
|
xinference/model/llm/lmdeploy/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
|
|
@@ -93,27 +94,27 @@ xinference/model/llm/sglang/__init__.py,sha256=-sjSIQ4K6w-TEzx49kVaWeWC443fnZqOD
|
|
|
93
94
|
xinference/model/llm/sglang/core.py,sha256=ft4QlDw36gwoic8lyjtSx2ai6KTW84CPVbYr8grqGMI,16698
|
|
94
95
|
xinference/model/llm/transformers/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQAL1ms,581
|
|
95
96
|
xinference/model/llm/transformers/chatglm.py,sha256=pCJgoc0Ikny5BL85iHXl9M6zKgIzAHBsd81NAZ31yRI,17834
|
|
96
|
-
xinference/model/llm/transformers/cogvlm2.py,sha256=
|
|
97
|
-
xinference/model/llm/transformers/cogvlm2_video.py,sha256=
|
|
97
|
+
xinference/model/llm/transformers/cogvlm2.py,sha256=I5Ftm0VYjbTAv5ZARZCo32Ggpw58PJfHs5B_nX_BIlU,15972
|
|
98
|
+
xinference/model/llm/transformers/cogvlm2_video.py,sha256=ZGkpC4x2uEtjwoMrLSODmAUYTjOeSNYxZi9VpQrpnhU,11857
|
|
98
99
|
xinference/model/llm/transformers/compression.py,sha256=U0vMJ-JaBt4oC2LffgWg6HbPj1CeUi_YdwVbjDd0mRA,8112
|
|
99
|
-
xinference/model/llm/transformers/core.py,sha256
|
|
100
|
+
xinference/model/llm/transformers/core.py,sha256=fyM7WXsVTEs38LAUPe-CJf45czInrNf1ifsWWexZ5nM,31128
|
|
100
101
|
xinference/model/llm/transformers/deepseek_v2.py,sha256=HSddUBm5sKpHTNtPbTao9r3Yif-_xRAJrAtfCyELnhw,12975
|
|
101
|
-
xinference/model/llm/transformers/deepseek_vl.py,sha256=
|
|
102
|
-
xinference/model/llm/transformers/glm4v.py,sha256=
|
|
103
|
-
xinference/model/llm/transformers/intern_vl.py,sha256=
|
|
102
|
+
xinference/model/llm/transformers/deepseek_vl.py,sha256=pB6i6DW5oyfHdqTgKpi2DkIKVGlPLGIDR_Op0sB1uKA,10445
|
|
103
|
+
xinference/model/llm/transformers/glm4v.py,sha256=goph2HhpV8gUm2t8-T1P-jTF2r_kPeH6QNe64lmlm0g,13871
|
|
104
|
+
xinference/model/llm/transformers/intern_vl.py,sha256=0pbze1eo3HvNQ0nW-mVJcJuJ4GrEyBBqQAYIdXnAn6c,18270
|
|
104
105
|
xinference/model/llm/transformers/internlm2.py,sha256=nRrmbH9bJl_wLcCH4zSy0EeKeP_ht-b8bVvbG2pMgV0,7228
|
|
105
|
-
xinference/model/llm/transformers/minicpmv25.py,sha256=
|
|
106
|
-
xinference/model/llm/transformers/minicpmv26.py,sha256=
|
|
107
|
-
xinference/model/llm/transformers/omnilmm.py,sha256=
|
|
108
|
-
xinference/model/llm/transformers/qwen2_audio.py,sha256=
|
|
109
|
-
xinference/model/llm/transformers/qwen2_vl.py,sha256=
|
|
110
|
-
xinference/model/llm/transformers/qwen_vl.py,sha256=
|
|
106
|
+
xinference/model/llm/transformers/minicpmv25.py,sha256=mr80-OlSlK_opSuAO3cz_QlkqujLr6V-OsTP0ebwpE8,6814
|
|
107
|
+
xinference/model/llm/transformers/minicpmv26.py,sha256=_e2C4vmyKIzKt7S7AvKgiqhDOhGiBXa6Xoiix4UaYtI,13440
|
|
108
|
+
xinference/model/llm/transformers/omnilmm.py,sha256=2ZLW979ETqDDKo9CaTNwi9uLBZ2d6itHAYqjUA4jdro,5172
|
|
109
|
+
xinference/model/llm/transformers/qwen2_audio.py,sha256=1XmlawVF-Xh2pgGoLDX7kOYIiF_bDUR3doSOnM59QbQ,6107
|
|
110
|
+
xinference/model/llm/transformers/qwen2_vl.py,sha256=i8mypQwaPaaGQ0OIS55H8yuUX6gH87ubPuPQHHAD9fw,7304
|
|
111
|
+
xinference/model/llm/transformers/qwen_vl.py,sha256=LG19qJW30bFiZOS-t9OM3JP6K1KCLj_Sv3nKSCLvyts,14058
|
|
111
112
|
xinference/model/llm/transformers/tensorizer_utils.py,sha256=VXSYbPZtCbd8lVvsnjDLPZjfCMil67Pkywd_Ze4dTx4,11362
|
|
112
|
-
xinference/model/llm/transformers/utils.py,sha256=
|
|
113
|
-
xinference/model/llm/transformers/yi_vl.py,sha256=
|
|
113
|
+
xinference/model/llm/transformers/utils.py,sha256=kTaNK65igHoWRUe00FD-Bs7nBv_OYre0KXjbmstlleU,29228
|
|
114
|
+
xinference/model/llm/transformers/yi_vl.py,sha256=iCdRLw-wizbU-qXXc8CT4DhC0Pt-uYg0vFwXEhAZjQg,8961
|
|
114
115
|
xinference/model/llm/vllm/__init__.py,sha256=h_JgzSqV5lP6vQ6XX_17kE4IY4BRnvKta_7VLQAL1ms,581
|
|
115
|
-
xinference/model/llm/vllm/core.py,sha256=
|
|
116
|
-
xinference/model/llm/vllm/utils.py,sha256=
|
|
116
|
+
xinference/model/llm/vllm/core.py,sha256=FhwRaRY29imMS4Aldda7qBQg4tCtUqG8adr1zfUF7jw,30729
|
|
117
|
+
xinference/model/llm/vllm/utils.py,sha256=LKOmwfFRrlSecawxT-uE39tC2RQbf1UIiSH9Uz90X6w,1313
|
|
117
118
|
xinference/model/rerank/__init__.py,sha256=wRpf1bOMfmAsuEKEGczMTB5fWEvuqltlJbIbRb-x8Ko,3483
|
|
118
119
|
xinference/model/rerank/core.py,sha256=ZOlbtK1x8CgcAAg0Y-5AF9ItYbhxuGtf7C_Sf0D9Kww,14122
|
|
119
120
|
xinference/model/rerank/custom.py,sha256=wPKF3bHbGap9dHz9yYvXMXhozh4hRzS78RQijqvaRq8,3846
|
|
@@ -121,7 +122,7 @@ xinference/model/rerank/model_spec.json,sha256=xUuJgJ8Ad4l2v8gEHxAdF_xoaSkA8j8AX
|
|
|
121
122
|
xinference/model/rerank/model_spec_modelscope.json,sha256=pf5hX4g0HdVjk2-ibHTl_mXHgQSSPYMTBOIwvnwMMkk,1616
|
|
122
123
|
xinference/model/rerank/utils.py,sha256=MJAFL47G3r3zLVGXKoi0QLTgU3Xr4Ffv72Ipn--psew,713
|
|
123
124
|
xinference/model/video/__init__.py,sha256=mRhOhzMxzcPFdA5j4niAxH_j9dXLtT9HmchuICrdET8,2160
|
|
124
|
-
xinference/model/video/core.py,sha256=
|
|
125
|
+
xinference/model/video/core.py,sha256=QEdVbVBDQebSWxqkL483Q2Y9Y1GGc2an0gi2QBPUH9I,6013
|
|
125
126
|
xinference/model/video/diffusers.py,sha256=kSEBRf0vtWyo0IrwoiEpr_ROu7SwDAVBZ4leqkcPycM,6244
|
|
126
127
|
xinference/model/video/model_spec.json,sha256=yQcLSU3vRJys-ACdHGtTNdz2pX1O9QDQ5rGHQd9LdFY,817
|
|
127
128
|
xinference/model/video/model_spec_modelscope.json,sha256=U8p6IqNLbY5Safxwpa6dCfnGbyvOC4FtYIf2ucr8TvM,815
|
|
@@ -15507,9 +15508,9 @@ xinference/web/ui/node_modules/yargs-parser/package.json,sha256=BSwbOzgetKXMK4u0
|
|
|
15507
15508
|
xinference/web/ui/node_modules/yocto-queue/package.json,sha256=6U1XHQPGXJTqsiFvT953ORihUtXTblZy4fXBWP9qxC0,725
|
|
15508
15509
|
xinference/web/ui/node_modules/yup/package.json,sha256=xRFSROB9NKxqSWHEVFvSTsPs9Ll074uo8OS1zEw0qhA,1206
|
|
15509
15510
|
xinference/web/ui/node_modules/yup/node_modules/type-fest/package.json,sha256=JTv2zTTVgxQ2H82m1-6qEpdMv08lHjFx4Puf_MsbB_Q,1134
|
|
15510
|
-
xinference-0.15.
|
|
15511
|
-
xinference-0.15.
|
|
15512
|
-
xinference-0.15.
|
|
15513
|
-
xinference-0.15.
|
|
15514
|
-
xinference-0.15.
|
|
15515
|
-
xinference-0.15.
|
|
15511
|
+
xinference-0.15.4.dist-info/LICENSE,sha256=QwcOLU5TJoTeUhuIXzhdCEEDDvorGiC6-3YTOl4TecE,11356
|
|
15512
|
+
xinference-0.15.4.dist-info/METADATA,sha256=PANLzbQqqFtpqMFBVCP9JcY30fhoC63e5dG3Y5hbbr4,19136
|
|
15513
|
+
xinference-0.15.4.dist-info/WHEEL,sha256=eOLhNAGa2EW3wWl_TU484h7q1UNgy0JXjjoqKoxAAQc,92
|
|
15514
|
+
xinference-0.15.4.dist-info/entry_points.txt,sha256=-lDyyzqWMFQF0Rgm7VxBNz0V-bMBMQLRR3pvQ-Y8XTY,226
|
|
15515
|
+
xinference-0.15.4.dist-info/top_level.txt,sha256=L1rQt7pl6m8tmKXpWVHzP-GtmzAxp663rXxGE7qnK00,11
|
|
15516
|
+
xinference-0.15.4.dist-info/RECORD,,
|
|
File without changes
|
|
File without changes
|
|
File without changes
|
|
File without changes
|