xinference 1.10.1__py3-none-any.whl → 1.11.0__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.

Files changed (38) hide show
  1. xinference/_version.py +3 -3
  2. xinference/api/restful_api.py +462 -3
  3. xinference/client/restful/async_restful_client.py +158 -5
  4. xinference/client/restful/restful_client.py +131 -0
  5. xinference/core/supervisor.py +12 -0
  6. xinference/model/audio/model_spec.json +20 -20
  7. xinference/model/image/model_spec.json +159 -159
  8. xinference/model/llm/__init__.py +2 -2
  9. xinference/model/llm/llm_family.json +843 -180
  10. xinference/model/llm/mlx/distributed_models/core.py +41 -0
  11. xinference/model/llm/mlx/distributed_models/qwen2.py +1 -2
  12. xinference/model/llm/sglang/core.py +20 -6
  13. xinference/model/llm/tool_parsers/qwen_tool_parser.py +29 -4
  14. xinference/model/llm/transformers/chatglm.py +3 -0
  15. xinference/model/llm/transformers/core.py +129 -36
  16. xinference/model/llm/transformers/multimodal/minicpmv45.py +340 -0
  17. xinference/model/llm/transformers/utils.py +23 -0
  18. xinference/model/llm/utils.py +37 -24
  19. xinference/model/llm/vllm/core.py +128 -69
  20. xinference/model/utils.py +74 -31
  21. xinference/thirdparty/audiotools/core/audio_signal.py +6 -6
  22. xinference/thirdparty/fish_speech/fish_speech/text/chn_text_norm/text.py +1 -1
  23. xinference/thirdparty/melo/text/chinese_mix.py +2 -2
  24. xinference/types.py +9 -0
  25. xinference/ui/web/ui/build/asset-manifest.json +3 -3
  26. xinference/ui/web/ui/build/index.html +1 -1
  27. xinference/ui/web/ui/build/static/js/{main.d192c4f3.js → main.45e78536.js} +3 -3
  28. xinference/ui/web/ui/build/static/js/main.45e78536.js.map +1 -0
  29. xinference/ui/web/ui/node_modules/.cache/babel-loader/ea2a26361204e70cf1018d6990fb6354bed82b3ac69690391e0f100385e7abb7.json +1 -0
  30. {xinference-1.10.1.dist-info → xinference-1.11.0.dist-info}/METADATA +7 -5
  31. {xinference-1.10.1.dist-info → xinference-1.11.0.dist-info}/RECORD +36 -35
  32. xinference/ui/web/ui/build/static/js/main.d192c4f3.js.map +0 -1
  33. xinference/ui/web/ui/node_modules/.cache/babel-loader/f995a2425dfb0822fd07127f66ffe9b026883bc156b402eb8bd0b83d52460a93.json +0 -1
  34. /xinference/ui/web/ui/build/static/js/{main.d192c4f3.js.LICENSE.txt → main.45e78536.js.LICENSE.txt} +0 -0
  35. {xinference-1.10.1.dist-info → xinference-1.11.0.dist-info}/WHEEL +0 -0
  36. {xinference-1.10.1.dist-info → xinference-1.11.0.dist-info}/entry_points.txt +0 -0
  37. {xinference-1.10.1.dist-info → xinference-1.11.0.dist-info}/licenses/LICENSE +0 -0
  38. {xinference-1.10.1.dist-info → xinference-1.11.0.dist-info}/top_level.txt +0 -0
xinference/model/image/model_spec.json CHANGED
@@ -254,8 +254,8 @@
254
254
  "torch_dtype": "bfloat16"
255
255
  },
256
256
  "default_generate_config": {
257
- "guidance_scale": 1.0,
258
- "true_cfg_scale": 1.0
257
+ "guidance_scale": 1,
258
+ "true_cfg_scale": 1
259
259
  },
260
260
  "virtualenv": {
261
261
  "packages": [
@@ -267,160 +267,6 @@
267
267
  "no_build_isolation": true
268
268
  }
269
269
  },
270
- {
271
- "version": 2,
272
- "model_name": "Qwen-Image-Edit",
273
- "model_family": "stable_diffusion",
274
- "model_ability": [
275
- "image2image"
276
- ],
277
- "model_src": {
278
- "huggingface": {
279
- "model_id": "Qwen/Qwen-Image-Edit",
280
- "model_revision": "0b71959872ea3bf4d106c578b7c480ebb133dba7",
281
- "gguf_model_id": "QuantStack/Qwen-Image-Edit-GGUF",
282
- "gguf_quantizations": [
283
- "Q2_K",
284
- "Q3_K_M",
285
- "Q3_K_S",
286
- "Q4_0",
287
- "Q4_1",
288
- "Q4_K_M",
289
- "Q4_K_S",
290
- "Q5_0",
291
- "Q5_1",
292
- "Q5_K_M",
293
- "Q5_K_S",
294
- "Q6_K",
295
- "Q8_0"
296
- ],
297
- "gguf_model_file_name_template": "Qwen_Image_Edit-{quantization}.gguf",
298
- "lightning_model_id": "lightx2v/Qwen-Image-Lightning",
299
- "lightning_versions": [
300
- "4steps-V1.0-bf16",
301
- "4steps-V1.0",
302
- "8steps-V1.0-bf16",
303
- "8steps-V1.0"
304
- ],
305
- "lightning_model_file_name_template": "Qwen-Image-Edit-Lightning-{lightning_version}.safetensors"
306
- },
307
- "modelscope": {
308
- "model_id": "Qwen/Qwen-Image-Edit",
309
- "model_revision": "master",
310
- "gguf_model_id": "QuantStack/Qwen-Image-Edit-GGUF",
311
- "gguf_quantizations": [
312
- "Q2_K",
313
- "Q3_K_M",
314
- "Q3_K_S",
315
- "Q4_0",
316
- "Q4_1",
317
- "Q4_K_M",
318
- "Q4_K_S",
319
- "Q5_0",
320
- "Q5_1",
321
- "Q5_K_M",
322
- "Q5_K_S",
323
- "Q6_K",
324
- "Q8_0"
325
- ],
326
- "gguf_model_file_name_template": "Qwen_Image_Edit-{quantization}.gguf",
327
- "lightning_model_id": "lightx2v/Qwen-Image-Lightning",
328
- "lightning_versions": [
329
- "4steps-V1.0-bf16",
330
- "4steps-V1.0",
331
- "8steps-V1.0-bf16",
332
- "8steps-V1.0"
333
- ],
334
- "lightning_model_file_name_template": "Qwen-Image-Edit-Lightning-{lightning_version}.safetensors"
335
- }
336
- },
337
- "default_model_config": {
338
- "quantize": true,
339
- "quantize_text_encoder": "text_encoder",
340
- "torch_dtype": "bfloat16"
341
- },
342
- "default_generate_config": {
343
- "true_cfg_scale": 4.0
344
- },
345
- "virtualenv": {
346
- "packages": [
347
- "diffusers==0.35.1",
348
- "peft>=0.17.0",
349
- "#system_torch#",
350
- "#system_numpy#"
351
- ],
352
- "no_build_isolation": true
353
- }
354
- },
355
- {
356
- "version": 2,
357
- "model_name": "Qwen-Image-Edit-2509",
358
- "model_family": "stable_diffusion",
359
- "model_ability": [
360
- "image2image"
361
- ],
362
- "model_src": {
363
- "huggingface": {
364
- "model_id": "Qwen/Qwen-Image-Edit-2509",
365
- "model_revision": "d3968ef930e841f4c73640fb8afa3b306a78167e",
366
- "gguf_model_id": "QuantStack/Qwen-Image-Edit-2509-GGUF",
367
- "gguf_quantizations": [
368
- "Q2_K",
369
- "Q3_K_M",
370
- "Q3_K_S",
371
- "Q4_0",
372
- "Q4_1",
373
- "Q4_K_M",
374
- "Q4_K_S",
375
- "Q5_0",
376
- "Q5_1",
377
- "Q5_K_M",
378
- "Q5_K_S",
379
- "Q6_K",
380
- "Q8_0"
381
- ],
382
- "gguf_model_file_name_template": "Qwen-Image-Edit-2509-{quantization}.gguf"
383
- },
384
- "modelscope": {
385
- "model_id": "Qwen/Qwen-Image-Edit-2509",
386
- "model_revision": "master",
387
- "gguf_model_id": "QuantStack/Qwen-Image-Edit-2509-GGUF",
388
- "gguf_quantizations": [
389
- "Q2_K",
390
- "Q3_K_M",
391
- "Q3_K_S",
392
- "Q4_0",
393
- "Q4_1",
394
- "Q4_K_M",
395
- "Q4_K_S",
396
- "Q5_0",
397
- "Q5_1",
398
- "Q5_K_M",
399
- "Q5_K_S",
400
- "Q6_K",
401
- "Q8_0"
402
- ],
403
- "gguf_model_file_name_template": "Qwen-Image-Edit-2509-{quantization}.gguf"
404
- }
405
- },
406
- "default_model_config": {
407
- "quantize": true,
408
- "quantize_text_encoder": "text_encoder",
409
- "torch_dtype": "bfloat16"
410
- },
411
- "default_generate_config": {
412
- "true_cfg_scale": 4.0
413
- },
414
- "virtualenv": {
415
- "packages": [
416
- "git+https://github.com/huggingface/diffusers",
417
- "peft>=0.17.0",
418
- "#system_torch#",
419
- "#system_numpy#"
420
- ],
421
- "no_build_isolation": true
422
- }
423
- },
424
270
  {
425
271
  "version": 2,
426
272
  "model_name": "sd3-medium",
@@ -566,7 +412,7 @@
566
412
  "transformer_quantization": "nf4"
567
413
  },
568
414
  "default_generate_config": {
569
- "guidance_scale": 1.0,
415
+ "guidance_scale": 1,
570
416
  "num_inference_steps": 4
571
417
  },
572
418
  "model_src": {
@@ -644,7 +490,7 @@
644
490
  "text2image"
645
491
  ],
646
492
  "default_generate_config": {
647
- "guidance_scale": 0.0,
493
+ "guidance_scale": 0,
648
494
  "num_inference_steps": 1
649
495
  },
650
496
  "model_src": {
@@ -666,7 +512,7 @@
666
512
  "text2image"
667
513
  ],
668
514
  "default_generate_config": {
669
- "guidance_scale": 0.0,
515
+ "guidance_scale": 0,
670
516
  "num_inference_steps": 1
671
517
  },
672
518
  "model_src": {
@@ -911,5 +757,159 @@
911
757
  "model_revision": "master"
912
758
  }
913
759
  }
760
+ },
761
+ {
762
+ "version": 2,
763
+ "model_name": "Qwen-Image-Edit",
764
+ "model_family": "stable_diffusion",
765
+ "model_ability": [
766
+ "image2image"
767
+ ],
768
+ "model_src": {
769
+ "huggingface": {
770
+ "model_id": "Qwen/Qwen-Image-Edit",
771
+ "model_revision": "0b71959872ea3bf4d106c578b7c480ebb133dba7",
772
+ "gguf_model_id": "QuantStack/Qwen-Image-Edit-GGUF",
773
+ "gguf_quantizations": [
774
+ "Q2_K",
775
+ "Q3_K_M",
776
+ "Q3_K_S",
777
+ "Q4_0",
778
+ "Q4_1",
779
+ "Q4_K_M",
780
+ "Q4_K_S",
781
+ "Q5_0",
782
+ "Q5_1",
783
+ "Q5_K_M",
784
+ "Q5_K_S",
785
+ "Q6_K",
786
+ "Q8_0"
787
+ ],
788
+ "gguf_model_file_name_template": "Qwen_Image_Edit-{quantization}.gguf",
789
+ "lightning_model_id": "lightx2v/Qwen-Image-Lightning",
790
+ "lightning_versions": [
791
+ "4steps-V1.0-bf16",
792
+ "4steps-V1.0",
793
+ "8steps-V1.0-bf16",
794
+ "8steps-V1.0"
795
+ ],
796
+ "lightning_model_file_name_template": "Qwen-Image-Edit-Lightning-{lightning_version}.safetensors"
797
+ },
798
+ "modelscope": {
799
+ "model_id": "Qwen/Qwen-Image-Edit",
800
+ "model_revision": "master",
801
+ "gguf_model_id": "QuantStack/Qwen-Image-Edit-GGUF",
802
+ "gguf_quantizations": [
803
+ "Q2_K",
804
+ "Q3_K_M",
805
+ "Q3_K_S",
806
+ "Q4_0",
807
+ "Q4_1",
808
+ "Q4_K_M",
809
+ "Q4_K_S",
810
+ "Q5_0",
811
+ "Q5_1",
812
+ "Q5_K_M",
813
+ "Q5_K_S",
814
+ "Q6_K",
815
+ "Q8_0"
816
+ ],
817
+ "gguf_model_file_name_template": "Qwen_Image_Edit-{quantization}.gguf",
818
+ "lightning_model_id": "lightx2v/Qwen-Image-Lightning",
819
+ "lightning_versions": [
820
+ "4steps-V1.0-bf16",
821
+ "4steps-V1.0",
822
+ "8steps-V1.0-bf16",
823
+ "8steps-V1.0"
824
+ ],
825
+ "lightning_model_file_name_template": "Qwen-Image-Edit-Lightning-{lightning_version}.safetensors"
826
+ }
827
+ },
828
+ "default_model_config": {
829
+ "quantize": true,
830
+ "quantize_text_encoder": "text_encoder",
831
+ "torch_dtype": "bfloat16"
832
+ },
833
+ "default_generate_config": {
834
+ "true_cfg_scale": 4
835
+ },
836
+ "virtualenv": {
837
+ "packages": [
838
+ "diffusers==0.35.1",
839
+ "peft>=0.17.0",
840
+ "#system_torch#",
841
+ "#system_numpy#"
842
+ ],
843
+ "no_build_isolation": true
844
+ }
845
+ },
846
+ {
847
+ "version": 2,
848
+ "model_name": "Qwen-Image-Edit-2509",
849
+ "model_family": "stable_diffusion",
850
+ "model_ability": [
851
+ "image2image"
852
+ ],
853
+ "model_src": {
854
+ "huggingface": {
855
+ "model_id": "Qwen/Qwen-Image-Edit-2509",
856
+ "model_revision": "d3968ef930e841f4c73640fb8afa3b306a78167e",
857
+ "gguf_model_id": "QuantStack/Qwen-Image-Edit-2509-GGUF",
858
+ "gguf_quantizations": [
859
+ "Q2_K",
860
+ "Q3_K_M",
861
+ "Q3_K_S",
862
+ "Q4_0",
863
+ "Q4_1",
864
+ "Q4_K_M",
865
+ "Q4_K_S",
866
+ "Q5_0",
867
+ "Q5_1",
868
+ "Q5_K_M",
869
+ "Q5_K_S",
870
+ "Q6_K",
871
+ "Q8_0"
872
+ ],
873
+ "gguf_model_file_name_template": "Qwen-Image-Edit-2509-{quantization}.gguf"
874
+ },
875
+ "modelscope": {
876
+ "model_id": "Qwen/Qwen-Image-Edit-2509",
877
+ "model_revision": "master",
878
+ "gguf_model_id": "QuantStack/Qwen-Image-Edit-2509-GGUF",
879
+ "gguf_quantizations": [
880
+ "Q2_K",
881
+ "Q3_K_M",
882
+ "Q3_K_S",
883
+ "Q4_0",
884
+ "Q4_1",
885
+ "Q4_K_M",
886
+ "Q4_K_S",
887
+ "Q5_0",
888
+ "Q5_1",
889
+ "Q5_K_M",
890
+ "Q5_K_S",
891
+ "Q6_K",
892
+ "Q8_0"
893
+ ],
894
+ "gguf_model_file_name_template": "Qwen-Image-Edit-2509-{quantization}.gguf"
895
+ }
896
+ },
897
+ "default_model_config": {
898
+ "quantize": true,
899
+ "quantize_text_encoder": "text_encoder",
900
+ "torch_dtype": "bfloat16"
901
+ },
902
+ "default_generate_config": {
903
+ "true_cfg_scale": 4
904
+ },
905
+ "virtualenv": {
906
+ "packages": [
907
+ "git+https://github.com/huggingface/diffusers",
908
+ "peft>=0.17.0",
909
+ "#system_torch#",
910
+ "#system_numpy#"
911
+ ],
912
+ "no_build_isolation": true
913
+ }
914
914
  }
915
915
  ]
xinference/model/llm/__init__.py CHANGED
@@ -179,12 +179,12 @@ def _install():
179
179
  from .mlx.core import MLXChatModel, MLXModel, MLXVisionModel
180
180
  from .sglang.core import SGLANGChatModel, SGLANGModel, SGLANGVisionModel
181
181
  from .transformers.core import PytorchChatModel, PytorchModel
182
- from .vllm.core import VLLMChatModel, VLLMModel, VLLMVisionModel
182
+ from .vllm.core import VLLMChatModel, VLLMModel, VLLMMultiModel
183
183
 
184
184
  # register llm classes.
185
185
  LLAMA_CLASSES.extend([XllamaCppModel])
186
186
  SGLANG_CLASSES.extend([SGLANGModel, SGLANGChatModel, SGLANGVisionModel])
187
- VLLM_CLASSES.extend([VLLMModel, VLLMChatModel, VLLMVisionModel])
187
+ VLLM_CLASSES.extend([VLLMModel, VLLMChatModel, VLLMMultiModel])
188
188
  MLX_CLASSES.extend([MLXModel, MLXChatModel, MLXVisionModel])
189
189
  LMDEPLOY_CLASSES.extend([LMDeployModel, LMDeployChatModel])
190
190
  TRANSFORMERS_CLASSES.extend([PytorchChatModel, PytorchModel])