sglang 0.2.14.post1__py3-none-any.whl → 0.2.15__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.
Files changed (63) hide show
  1. sglang/api.py +2 -0
  2. sglang/bench_latency.py +39 -28
  3. sglang/lang/interpreter.py +3 -0
  4. sglang/lang/ir.py +5 -0
  5. sglang/launch_server_llavavid.py +26 -0
  6. sglang/srt/configs/__init__.py +5 -0
  7. sglang/srt/configs/exaone.py +195 -0
  8. sglang/srt/constrained/fsm_cache.py +1 -1
  9. sglang/srt/conversation.py +24 -2
  10. sglang/srt/hf_transformers_utils.py +11 -160
  11. sglang/srt/layers/activation.py +10 -4
  12. sglang/srt/layers/extend_attention.py +13 -8
  13. sglang/srt/layers/layernorm.py +47 -1
  14. sglang/srt/layers/logits_processor.py +4 -4
  15. sglang/srt/layers/sampler.py +69 -16
  16. sglang/srt/managers/controller_multi.py +5 -5
  17. sglang/srt/managers/controller_single.py +5 -5
  18. sglang/srt/managers/io_struct.py +11 -5
  19. sglang/srt/managers/schedule_batch.py +25 -13
  20. sglang/srt/managers/tokenizer_manager.py +76 -63
  21. sglang/srt/managers/tp_worker.py +47 -36
  22. sglang/srt/model_config.py +3 -3
  23. sglang/srt/model_executor/cuda_graph_runner.py +24 -9
  24. sglang/srt/model_executor/forward_batch_info.py +78 -43
  25. sglang/srt/model_executor/model_runner.py +29 -18
  26. sglang/srt/models/chatglm.py +5 -13
  27. sglang/srt/models/commandr.py +5 -1
  28. sglang/srt/models/dbrx.py +5 -1
  29. sglang/srt/models/deepseek.py +5 -1
  30. sglang/srt/models/deepseek_v2.py +57 -25
  31. sglang/srt/models/exaone.py +399 -0
  32. sglang/srt/models/gemma.py +7 -3
  33. sglang/srt/models/gemma2.py +6 -52
  34. sglang/srt/models/gpt_bigcode.py +5 -1
  35. sglang/srt/models/grok.py +14 -4
  36. sglang/srt/models/internlm2.py +5 -1
  37. sglang/srt/models/llama2.py +10 -7
  38. sglang/srt/models/llama_classification.py +2 -6
  39. sglang/srt/models/llama_embedding.py +3 -4
  40. sglang/srt/models/llava.py +69 -91
  41. sglang/srt/models/llavavid.py +40 -86
  42. sglang/srt/models/minicpm.py +5 -1
  43. sglang/srt/models/mixtral.py +6 -2
  44. sglang/srt/models/mixtral_quant.py +5 -1
  45. sglang/srt/models/qwen.py +5 -2
  46. sglang/srt/models/qwen2.py +9 -6
  47. sglang/srt/models/qwen2_moe.py +12 -33
  48. sglang/srt/models/stablelm.py +5 -1
  49. sglang/srt/models/yivl.py +2 -7
  50. sglang/srt/openai_api/adapter.py +16 -1
  51. sglang/srt/openai_api/protocol.py +5 -5
  52. sglang/srt/sampling/sampling_batch_info.py +79 -6
  53. sglang/srt/server.py +9 -9
  54. sglang/srt/utils.py +18 -36
  55. sglang/test/runners.py +2 -2
  56. sglang/test/test_layernorm.py +53 -1
  57. sglang/version.py +1 -1
  58. {sglang-0.2.14.post1.dist-info → sglang-0.2.15.dist-info}/METADATA +8 -8
  59. sglang-0.2.15.dist-info/RECORD +118 -0
  60. sglang-0.2.14.post1.dist-info/RECORD +0 -114
  61. {sglang-0.2.14.post1.dist-info → sglang-0.2.15.dist-info}/LICENSE +0 -0
  62. {sglang-0.2.14.post1.dist-info → sglang-0.2.15.dist-info}/WHEEL +0 -0
  63. {sglang-0.2.14.post1.dist-info → sglang-0.2.15.dist-info}/top_level.txt +0 -0
@@ -1,114 +0,0 @@
1
- sglang/__init__.py,sha256=T8MYdFfKFPZcgFKHMBpOCIlFbhjwmr77Nqm6mdE6bCY,1590
2
- sglang/api.py,sha256=8B_ADgLN2fjo9Ej123hInfHA4wmpUkV0yyErSiRnfAA,6408
3
- sglang/bench_latency.py,sha256=VEdGBX5vZSngS8AeOdJJRW65BIJsZXhKwAK5z20SZoI,16344
4
- sglang/bench_serving.py,sha256=J_mMwnmDn0Jt07mzdGAuYOxpockHPLYJFL-kwoaqASY,36527
5
- sglang/check_env.py,sha256=rGRABCgt-0SfUrow4px28b2P59aMn8eVTnN5eZc_a8s,5397
6
- sglang/global_config.py,sha256=nwOjUflwqLQySPUMvk8Hk63TIS6mknh_ODSW3CZ1rJw,1704
7
- sglang/launch_server.py,sha256=FODfO0DW546dh-u1qDlWtrhsmj6hxkarXXv3cIdgkj8,549
8
- sglang/utils.py,sha256=zFYGkC4vOUR3sTv1TmQXcsOLZDtDBR3wnjqnDp3xMIs,8352
9
- sglang/version.py,sha256=-MR2TfhYp1Y9iHDXSv_cRGYS5Mzv-azCfIoLfZjKpgg,29
10
- sglang/lang/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
11
- sglang/lang/chat_template.py,sha256=uqI_I9zIKXGXg7-W-yjqvx1ZeS_TuwFCms6wkmC2QmY,13411
12
- sglang/lang/choices.py,sha256=-W1DVw9N9ZliVpvmWrzIXG4cswAah8eMQrHWzkS3D8o,6234
13
- sglang/lang/compiler.py,sha256=o1C6G3TzhjSlsH-doTPy5oiVehr57dxNTa5oZw5TTAI,7639
14
- sglang/lang/interpreter.py,sha256=-9VjAb5JqlxtBuQUDT08Cj2BW8VbLxTmJACe2cqza-s,30215
15
- sglang/lang/ir.py,sha256=GRcPsEjnR4k5q5Kf-Rb2YgDBseCTGQoasclhjmQtL8Y,17511
16
- sglang/lang/tracer.py,sha256=borJmlSJOhg1RUndGRnilnR60eEZz2Y9aU7BpftsOxU,8287
17
- sglang/lang/backend/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
18
- sglang/lang/backend/anthropic.py,sha256=EXRX7xJgA5KZszX7toSLVnKzFQ5EO0Loj-YjHFtxSxg,2081
19
- sglang/lang/backend/base_backend.py,sha256=Q5HdiDtyBewQeoYH0kDtBRVL8KFiEPNq9dw7XmauHQ8,1985
20
- sglang/lang/backend/litellm.py,sha256=ugmL7sfUxkUHVbHtwNzHgdQAEd4UCjNQboFuE3KThcY,2450
21
- sglang/lang/backend/openai.py,sha256=qM7eVH_kMxnDd2rpxOH0v76KxtOJFlAwgLgWIKvFGCI,15060
22
- sglang/lang/backend/runtime_endpoint.py,sha256=SDlp03EuQEK1eGK4_IaFySWgxlp4wCs3EPewZ6O640E,9549
23
- sglang/lang/backend/vertexai.py,sha256=O-iBLD-y3vq80UxnrAoJri7bxpgd-_eakZ88Cf8bEGA,4855
24
- sglang/srt/conversation.py,sha256=Ze2_dTHG6jc04ti7vuOEnoEe1ehvhxCJRpa4EYD0T_8,18494
25
- sglang/srt/hf_transformers_utils.py,sha256=OP5uBwnWiam6h9QvkBaG-nrDgkEUEwLXy1IWvW7rrRo,11737
26
- sglang/srt/mm_utils.py,sha256=zox644S3IHUWmADdK4MnIbdTS2DWHOy0_Dq0gCU38QQ,12273
27
- sglang/srt/model_config.py,sha256=k4OfRV-szWkFaJMIC40JoJGJ75AfYQ2hf4M1dS1aQ-o,6366
28
- sglang/srt/server.py,sha256=UE2Vc7abITVe-yhe9WpnCCa_FdeDPUY8462uSpGDutA,19807
29
- sglang/srt/server_args.py,sha256=GiDyPWCvYA_98mSE9LuvUoEodo9gRnNPPIPn0nFkxUs,18259
30
- sglang/srt/utils.py,sha256=x9MdBu0e8HAgaNIGuxiMVL7_nh03kl_rWuMnLas_Dgo,24327
31
- sglang/srt/constrained/__init__.py,sha256=NLpZGj9RIx83ejDrM_pfaRtqGgaPq_ggJszPQENUJ2E,2037
32
- sglang/srt/constrained/base_tool_cache.py,sha256=5sazBMHHDpHMoqOjuY6itCxwTmIFCflIWEDXMtmrPVs,2006
33
- sglang/srt/constrained/fsm_cache.py,sha256=1gvJzRxt6N_4KM1phlrFGPBquYiOUs_d8XIc7wztVk8,3100
34
- sglang/srt/constrained/jump_forward.py,sha256=LWRsmGPQcH6KT87wXwCRqtblU3pcAVCEzO0nWPxevs0,6636
35
- sglang/srt/layers/activation.py,sha256=CSKmMI2vLwLZL3heWdkbd3vYWzerIdmlvKJGKdu5kgk,4581
36
- sglang/srt/layers/decode_attention.py,sha256=TPD_608ZX9fQ_HDImifkxG_qcEYmimbEYY8lCBIjFuM,16628
37
- sglang/srt/layers/extend_attention.py,sha256=h4O0R7PJpAVKS3Vx_583zhrFPD0vv6XqzvOcHBI3zoc,14268
38
- sglang/srt/layers/layernorm.py,sha256=RzN4eESN9S8mw32r2Nxarq7wKFdeG1yhxPmehUMx79s,2073
39
- sglang/srt/layers/logits_processor.py,sha256=1hrxgvV1_qDsHiDak1JH6Vhn-hmkzi3AberLQ6-qkWw,13081
40
- sglang/srt/layers/pooler.py,sha256=qNMG3Ycvt2yf9mk1Lcs-2K7oPeCuVeDYoHAxkMu9b_Q,1610
41
- sglang/srt/layers/prefill_attention.py,sha256=y7vdcuX8lMa9Qf_jQYNDvQO9PVCBQSs3hb5LV2DFgpU,5256
42
- sglang/srt/layers/radix_attention.py,sha256=o5a8r3XQ-oRwaxBlAgzJGv7p3dMbu0LrYsDc4uvpPgA,8338
43
- sglang/srt/layers/sampler.py,sha256=XC6JcpuE1jO794u9AGRgzQ8xuu99G0l8PZKbX3_he7o,3995
44
- sglang/srt/layers/fused_moe/__init__.py,sha256=bWCrDdOy2ANEXTb8CHYO63O3Iu3eZnn0PJbgl0z5vvE,75
45
- sglang/srt/layers/fused_moe/fused_moe.py,sha256=1WM2cObWXcFWtqh_utGJFPnrT344rORwuQ9hJDaH2s0,23104
46
- sglang/srt/layers/fused_moe/layer.py,sha256=GT3r2UPx_PAufJd0SUMOXyh76ymAeYDubd0SM0H71bo,20977
47
- sglang/srt/managers/controller_multi.py,sha256=R45ST6oBlIwfUwuibMw0sgTk8iqphb_rFyIdW048JA4,6472
48
- sglang/srt/managers/controller_single.py,sha256=tnc71OTe8KDYouMdfqgwBT4lX5nZt6Rak9t2GmKtAME,5119
49
- sglang/srt/managers/detokenizer_manager.py,sha256=yQkL5gLomLiy1qc6e9HNz8hcj7JQFHm1AfIrzpXaWJE,6852
50
- sglang/srt/managers/io_struct.py,sha256=4Cs655K4n_F_usu6R3YE5_RdcE0XO9AXQNk5vl2II2c,10534
51
- sglang/srt/managers/policy_scheduler.py,sha256=7HNUxBKJE444s_bHcPpbnHCygsnH-NIXYNSC2q6mRmc,8584
52
- sglang/srt/managers/schedule_batch.py,sha256=hkCf0IGxZanXLMSwfg5lub0NFNRGuykumAUtCPeryBU,25578
53
- sglang/srt/managers/tokenizer_manager.py,sha256=aaZV7G3-m35pba1meRapqO7bdPjM2Cmkue5lbR_Jv3M,28836
54
- sglang/srt/managers/tp_worker.py,sha256=-icuKdRrfb51hx8BsmQUlBQVpsvWlZ34Z2qIdPZSLrM,36433
55
- sglang/srt/mem_cache/base_prefix_cache.py,sha256=qEQwEkG4E5rab2ZoTqcesf5pR_J4nV2jBxIHsBJHtIM,924
56
- sglang/srt/mem_cache/chunk_cache.py,sha256=CjZZYlqQzq7mYOiBMLWA5XNb6HIyh5lIMdY-K0OUZEc,2368
57
- sglang/srt/mem_cache/flush_cache.py,sha256=pTLKPRB17U6vl5RFJJvuJ4jCL2SyomgkUBNlkDpGRqo,978
58
- sglang/srt/mem_cache/memory_pool.py,sha256=4br3Ea2bfA-YsF_sPOVHlF2zQzYGd8fVaYTp197yZsE,7871
59
- sglang/srt/mem_cache/radix_cache.py,sha256=0AVr1BKKDOtTyybUkwxrz6PT8khDx-DpzgN5MgL27IE,10088
60
- sglang/srt/model_executor/cuda_graph_runner.py,sha256=X4sC6xCnjsYjaM8YStyu2_IDIMbD1Muntd5HO1W-wiw,12011
61
- sglang/srt/model_executor/forward_batch_info.py,sha256=Pv4vMXvCyJzRku2EHSVaNng28qglZQeska9wVkKUhyE,15544
62
- sglang/srt/model_executor/model_runner.py,sha256=rXyGUQaTZDyi_AU3QLY5mRIf-h0yzJrtJ1PtnCuyxNM,24064
63
- sglang/srt/models/chatglm.py,sha256=aoEgA2nflcOCIKtZojhUoboqxSP6i5IrrvuDOpzNPnE,13844
64
- sglang/srt/models/commandr.py,sha256=2rAXRZRb4PkJZ4NWEqP_rIgsjxbdZyHpuoMOarqTWzQ,14163
65
- sglang/srt/models/dbrx.py,sha256=N_0Ku_p1NCsc29NktUBNqPv7Z33XhYxOZK5xN7nzW4s,14661
66
- sglang/srt/models/deepseek.py,sha256=7UJgde1EV9ey6d-CKRcEyTKh1_WhZdatpZiltIuqpik,16006
67
- sglang/srt/models/deepseek_v2.py,sha256=uk--2a1e83H6U9wTx_wd3UvkS3VrSRSkjCOjky0R0uo,27004
68
- sglang/srt/models/gemma.py,sha256=3orOUznoGt2NxVKO5c8AjD_ue0gWqwb7LnKbhlcS5Vg,12276
69
- sglang/srt/models/gemma2.py,sha256=z-zhvLy3CX6OA_LF-lJSjpbN6dlyqFrHG4Gai7l0eM0,16584
70
- sglang/srt/models/gpt_bigcode.py,sha256=kzHYogeGXZF4KHpkXA-RGqvs016mA-6klWxD2QJTi9E,10195
71
- sglang/srt/models/grok.py,sha256=mdzd-fz8FirICdcpYLD94vwY3_ca0lxpIze8fAqDbHg,14805
72
- sglang/srt/models/internlm2.py,sha256=6j7JH0p3yib8GZDH8Cmrs-pgwfH3eOlAK6V3Cq64O7w,12202
73
- sglang/srt/models/llama2.py,sha256=HmzE1I8OnesmrdPY5b56l7okhWH_lRvWAg16K-UwKHg,14300
74
- sglang/srt/models/llama_classification.py,sha256=Dvzy3PfETiJtnKFOk8qDDLUoZECf_cpSrNeA60PaDo4,4932
75
- sglang/srt/models/llama_embedding.py,sha256=NQCQ3MnK3iRohL-UdY5UWxW4LlZ3RQZ7w4mlFOnpVrM,3696
76
- sglang/srt/models/llava.py,sha256=iuXLJVDWBiYo8zJuDPSSjt2LYqbkg2MAcOFUZO1fOX4,24353
77
- sglang/srt/models/llavavid.py,sha256=MX7YpqYh5J4BoOnV7vVAIfoOlBFQXYpp8Kpe7WK0ejk,13562
78
- sglang/srt/models/minicpm.py,sha256=ioqCsTCE_oF8xqGF5fm5cK9dclK5Y0EQ1UJfyteIDDo,13825
79
- sglang/srt/models/mistral.py,sha256=jlrWBVNXbAUziAaIdHAjFcOJnKtn9Bl8rBd65ypJM-I,819
80
- sglang/srt/models/mixtral.py,sha256=cZK-1kGXQC8ZC0tFNmbAoqWlyrrvv5omumpDdEwzzss,13623
81
- sglang/srt/models/mixtral_quant.py,sha256=wMACJq78OTWj7HlqPDRNEh8cjrVAjKqJEsOG3CO5xow,14072
82
- sglang/srt/models/qwen.py,sha256=ssdSgVuhT1Ei0JPa0xwqzrwwPNwkCHRJA4q70hK-Z7E,9988
83
- sglang/srt/models/qwen2.py,sha256=RHzzHfrq3_VoYkjoJu10z0PtyQ_Mit3mHHAO07214q0,12577
84
- sglang/srt/models/qwen2_moe.py,sha256=-Ijn_H2IGCjQAYA-9teS9IXKTPMBWSkkPp0Nox6MCuQ,17729
85
- sglang/srt/models/stablelm.py,sha256=30ngpc0Xq3VxzXJlf6svP1oax8Q3krMJkxM8PVKtZWU,11359
86
- sglang/srt/models/yivl.py,sha256=p4s_D_m4H2exP4b91Y-CTkq8T-eIG3DJsFy9pB0e7TM,4932
87
- sglang/srt/openai_api/adapter.py,sha256=5k2YGX6rEwpYlF3LaDs5Qvh6vD25LhAnPKWhcj2vnT0,49385
88
- sglang/srt/openai_api/protocol.py,sha256=73i1iFVI0tAJCLMV9QT4rgMzbyYoZ8-iuTIYs_Yx0Xo,9487
89
- sglang/srt/sampling/sampling_batch_info.py,sha256=TJLKc-FXqrI3-hHKEf-Bfu-W8NxNOEkmNvPaTN1O-g4,4911
90
- sglang/srt/sampling/sampling_params.py,sha256=ggOXxafqfCD-xrGYcM57byLZ79CIeBP4AD5F44L_CW0,5635
91
- sglang/srt/sampling/penaltylib/__init__.py,sha256=5vQw0Y5DSzmsoFg1IdMIKLwFVhYZ5ArADHVBYbSmOec,513
92
- sglang/srt/sampling/penaltylib/orchestrator.py,sha256=WkTNeDhj9H9rtp2ZZeX6MS2sdKSGlLboE6FcuKrwUo0,10815
93
- sglang/srt/sampling/penaltylib/penalizers/frequency_penalty.py,sha256=IvYioX53Vq_ji-0Zhcz_r5mUa3T3GaIydVS6K4FhWfE,2557
94
- sglang/srt/sampling/penaltylib/penalizers/min_new_tokens.py,sha256=XJZP0C4NFyXgcODbIWXxrgVEjmRgqLdZuVAtoN-LveY,3565
95
- sglang/srt/sampling/penaltylib/penalizers/presence_penalty.py,sha256=0PlANTrR959foTA3Nj5qBE7ndaOZgG-9X6LhzlmEUc8,2533
96
- sglang/srt/sampling/penaltylib/penalizers/repetition_penalty.py,sha256=v9jOgA0-I31WcrhIydiFbpy2ZJPLytFLGM98NRPd2sU,2820
97
- sglang/test/run_eval.py,sha256=NWxeLWmInBgkCvC9Jr_QzF7GfAiBve3Gf1JQrEOlNlU,3899
98
- sglang/test/runners.py,sha256=7bh1ApkK6LCouQJzFXi4s2woFvmIUkh6wDeQA2luGCY,7686
99
- sglang/test/simple_eval_common.py,sha256=r0G-9QLycs2ax3RMc44T_61fzMxlpTzv6pececC7lyY,12379
100
- sglang/test/simple_eval_gpqa.py,sha256=8Xt9Bw05c7SZTYrCZgB68OZUqUbLo69ywiyx0bTvSUk,3220
101
- sglang/test/simple_eval_humaneval.py,sha256=7lTi841NT58smNOtRwCedrdX9IWWypdLkOtaQOBy-GI,5687
102
- sglang/test/simple_eval_math.py,sha256=6kGKNwNbLN-Af3Wj8WTimWhH-Xp3enDmSvvSjsgWUpk,2550
103
- sglang/test/simple_eval_mgsm.py,sha256=wfbqJW9Rkc66vzq2fEMF6jchmoA8mw1OUiGU55cZ2B0,10261
104
- sglang/test/simple_eval_mmlu.py,sha256=FkwamjGMjueTixymkedF-YiPloSLiy4ftILFUrKZ9XI,4357
105
- sglang/test/test_activation.py,sha256=jkdNRzJnbd5OgZliQaIXpxovlcky17UrweomcOcMxoE,1442
106
- sglang/test/test_layernorm.py,sha256=VDdoeqGvebUa-l3rDiid6cC7wZq0Phpbm5fxxD0-cpg,1910
107
- sglang/test/test_programs.py,sha256=V_-Bx3lLkw37P6gDyA7mZCqxlyNMaFLBkRrPMQQQqn4,14909
108
- sglang/test/test_utils.py,sha256=HD-9rcj7EFS_NX1GQFU5613ITQlZaTK2l9RmqA0F7x4,14380
109
- sglang/test/srt/sampling/penaltylib/utils.py,sha256=-0p0rV-P4lNo7xAe3rQSBHTubc50a-DFyOQmLGAkgkQ,12515
110
- sglang-0.2.14.post1.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
111
- sglang-0.2.14.post1.dist-info/METADATA,sha256=FstugZ3Zw82h3E_xf6fNBN9HjdMPF4kyXiQA_KpQ3ak,37163
112
- sglang-0.2.14.post1.dist-info/WHEEL,sha256=UvcQYKBHoFqaQd6LKyqHw9fxEolWLQnlzP0h_LgJAfI,91
113
- sglang-0.2.14.post1.dist-info/top_level.txt,sha256=yxhh3pYQkcnA7v3Bg889C2jZhvtJdEincysO7PEB09M,7
114
- sglang-0.2.14.post1.dist-info/RECORD,,