sglang 0.3.4.post1__py3-none-any.whl → 0.3.4.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.
Files changed (30) hide show
  1. sglang/srt/configs/model_config.py +25 -2
  2. sglang/srt/constrained/fsm_cache.py +10 -3
  3. sglang/srt/hf_transformers_utils.py +14 -0
  4. sglang/srt/layers/attention/flashinfer_backend.py +5 -5
  5. sglang/srt/layers/logits_processor.py +5 -5
  6. sglang/srt/layers/rotary_embedding.py +15 -48
  7. sglang/srt/layers/sampler.py +51 -39
  8. sglang/srt/managers/data_parallel_controller.py +1 -1
  9. sglang/srt/managers/detokenizer_manager.py +4 -0
  10. sglang/srt/managers/io_struct.py +10 -0
  11. sglang/srt/managers/schedule_batch.py +13 -3
  12. sglang/srt/managers/scheduler.py +8 -2
  13. sglang/srt/managers/tokenizer_manager.py +14 -0
  14. sglang/srt/managers/tp_worker_overlap_thread.py +58 -21
  15. sglang/srt/mem_cache/memory_pool.py +10 -3
  16. sglang/srt/model_executor/cuda_graph_runner.py +29 -21
  17. sglang/srt/model_executor/forward_batch_info.py +6 -9
  18. sglang/srt/model_executor/model_runner.py +2 -2
  19. sglang/srt/sampling/penaltylib/penalizers/min_new_tokens.py +6 -3
  20. sglang/srt/sampling/sampling_params.py +5 -7
  21. sglang/srt/server.py +12 -0
  22. sglang/test/run_eval.py +2 -0
  23. sglang/test/srt/sampling/penaltylib/utils.py +1 -0
  24. sglang/test/test_utils.py +100 -3
  25. sglang/version.py +1 -1
  26. {sglang-0.3.4.post1.dist-info → sglang-0.3.4.post2.dist-info}/METADATA +13 -14
  27. {sglang-0.3.4.post1.dist-info → sglang-0.3.4.post2.dist-info}/RECORD +30 -30
  28. {sglang-0.3.4.post1.dist-info → sglang-0.3.4.post2.dist-info}/LICENSE +0 -0
  29. {sglang-0.3.4.post1.dist-info → sglang-0.3.4.post2.dist-info}/WHEEL +0 -0
  30. {sglang-0.3.4.post1.dist-info → sglang-0.3.4.post2.dist-info}/top_level.txt +0 -0
@@ -8,7 +8,7 @@ sglang/global_config.py,sha256=1r_W9rrBxGCCc2eqESRduOMMNq46e54xLgFLifHuQm0,1014
8
8
  sglang/launch_server.py,sha256=UnjNjYuZ8TtvmRtgYEsFImkbvCwvn_tQjk0V7cHy67E,450
9
9
  sglang/launch_server_llavavid.py,sha256=tGc17S1vUfLwbi1GB26oOdXxTWr7gjlqpTrPnrMRNO8,1007
10
10
  sglang/utils.py,sha256=NA_4xUrTI7KICQ3PEACfNWKE3nxSA5QvQZJNd4TQrDc,9395
11
- sglang/version.py,sha256=pqpAGctmlDC8XHDMyjIts8rLrvifYvVonBTzhD7nBfI,28
11
+ sglang/version.py,sha256=XbWUUFDK_Ls5pw5MPNOhsEKwI18aKVrvco5r4BTbEgU,28
12
12
  sglang/lang/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
13
13
  sglang/lang/chat_template.py,sha256=pmSCliE2vLCwx6m588Nv6QybNPc_o_CzSDS_YFmCmEo,14051
14
14
  sglang/lang/choices.py,sha256=-W1DVw9N9ZliVpvmWrzIXG4cswAah8eMQrHWzkS3D8o,6234
@@ -24,31 +24,31 @@ sglang/lang/backend/openai.py,sha256=qM7eVH_kMxnDd2rpxOH0v76KxtOJFlAwgLgWIKvFGCI
24
24
  sglang/lang/backend/runtime_endpoint.py,sha256=iVb7SlrpJ1ic92QG5kQUphZUb2EaVWY43dkmAO5pju4,10514
25
25
  sglang/lang/backend/vertexai.py,sha256=O-iBLD-y3vq80UxnrAoJri7bxpgd-_eakZ88Cf8bEGA,4855
26
26
  sglang/srt/conversation.py,sha256=erz6wEXMcSmBlskuUhX2c-MT0EMyqyFpTem9PgastEE,21107
27
- sglang/srt/hf_transformers_utils.py,sha256=viAO8ARv6efpasXJx8ZcNptivMs2MVENPTG9tZ6CwBU,6128
27
+ sglang/srt/hf_transformers_utils.py,sha256=3G41c2HbEsI-aN6VtbSo4MyqVhM5ZtTTiRM9NQ6jMzQ,6597
28
28
  sglang/srt/mm_utils.py,sha256=zox644S3IHUWmADdK4MnIbdTS2DWHOy0_Dq0gCU38QQ,12273
29
- sglang/srt/server.py,sha256=65b39k4FN_TzL8qAimS1mRx8xdO8jmKCdUftOISUv7M,26809
29
+ sglang/srt/server.py,sha256=uagMRPk-9n2mpOXbLhjiBSK3nGqGwh3ZyzKu8ZnsDUg,27204
30
30
  sglang/srt/server_args.py,sha256=AQHVnYJs62FBCFAvq4KYZtdvqGi0iOr41aFuHpb2qQw,27813
31
31
  sglang/srt/utils.py,sha256=t7GqQXzgGqJ_9v0_gEgr8v4_6zf86OHXthEvVe13Rsw,23710
32
32
  sglang/srt/configs/__init__.py,sha256=_usVIXHQjft4PAJ1Y-yGQOn2QNOv501GYMlQwpGXbns,208
33
33
  sglang/srt/configs/exaone.py,sha256=Duxd4yQoKy8GWEzZD_kCY_OzmN_67CTJL_Kgn0eXk3g,10731
34
- sglang/srt/configs/model_config.py,sha256=yqH3RGD3VNt5OU9V17EYe1Rx-AWCErJz_G3uqGJhjVc,6798
34
+ sglang/srt/configs/model_config.py,sha256=gSRANcpoB9o5je9a_atYokIo9CsRBJhAtg8FmZbLmSo,8026
35
35
  sglang/srt/configs/qwen2vl.py,sha256=AYHuFgJ0bwhWYkD7S6fvP7yJejJnuhy4xp5Q2W-O6ps,4424
36
36
  sglang/srt/constrained/__init__.py,sha256=ze8awDPvwAzdeMwzJ-25kXOQ4nVWoaP55jBDt5UOS_4,2070
37
37
  sglang/srt/constrained/base_tool_cache.py,sha256=5sazBMHHDpHMoqOjuY6itCxwTmIFCflIWEDXMtmrPVs,2006
38
- sglang/srt/constrained/fsm_cache.py,sha256=9GtliIN55Ov8Q9MSFfQC5rKrz3qTsB7Cm5OkhivKngY,3271
38
+ sglang/srt/constrained/fsm_cache.py,sha256=CWwtOHTU3sHmw71OhWxl05YgU7cNNVWohlUt71rG230,3536
39
39
  sglang/srt/constrained/jump_forward.py,sha256=o-CzJu3DEs0eFKlLzsQVYMSo4vBKpffs25sXLOJd6jc,6997
40
40
  sglang/srt/layers/activation.py,sha256=7VEkCrx2dvl629Lz0fkJcJfVoZA-ykEdkpTzKEc_drQ,5225
41
41
  sglang/srt/layers/layernorm.py,sha256=HCj8Y_X6MNNdtQU2sWKgyjIqVERxl9dqrmjbBbyJjpE,3796
42
42
  sglang/srt/layers/linear.py,sha256=EOdlpAf6srqxzvPpxcv10KFJKedNc22CGP1qEvpRbDg,46131
43
- sglang/srt/layers/logits_processor.py,sha256=Fq7VHwjP4iSzl_OBLo8qw_HVbIDbYB-0MGmfiD3Jk_E,12521
43
+ sglang/srt/layers/logits_processor.py,sha256=1l-hJoeZUfrPPmCWcyscl0ThgKWpprUELiL1mVDfbPE,12556
44
44
  sglang/srt/layers/pooler.py,sha256=rj2lygvleBnyLCBZ8I11HGMgpfIDsT0l3PIkshJwdu4,1606
45
45
  sglang/srt/layers/radix_attention.py,sha256=i07VRXPDHj-zJ1TSrXEqCxumQwYSHwAvc8DoIg-Irtg,1964
46
- sglang/srt/layers/rotary_embedding.py,sha256=iFVZw-jmUP7kJ3RHJzxHISPtGLhs-aPua7xTk8pxCWE,5274
47
- sglang/srt/layers/sampler.py,sha256=23wRDw2Fs3wZfPBh6gFBz2vqwxnoDd9LAHWq7YdQWlc,4166
46
+ sglang/srt/layers/rotary_embedding.py,sha256=gfRKBB8FmsQKiDH0Crh_KRIGRUuvEgazH1p_n9D_m7E,3889
47
+ sglang/srt/layers/sampler.py,sha256=3zfth1Kz24X4sUq7Z_cjZwHgPVivI-rgPtIeUbsiiWU,4589
48
48
  sglang/srt/layers/torchao_utils.py,sha256=1nzZkSzbF4qCAMeBKAeeDpMl_mK8imiY2RL3xFEgvAw,3340
49
49
  sglang/srt/layers/attention/__init__.py,sha256=EL1o6Q5vLgViN3pOr2A7F6K9FlNEpMdBypFAVMeq_HA,2445
50
50
  sglang/srt/layers/attention/double_sparsity_backend.py,sha256=BlX7uXteQpnoOnKsdBKh8h20zMVMEiibB5F_PkZSlNI,10706
51
- sglang/srt/layers/attention/flashinfer_backend.py,sha256=EgXOB6rAobB9g-EqNqP0tZo_atNEC5jOWQt9a7bHPtY,23280
51
+ sglang/srt/layers/attention/flashinfer_backend.py,sha256=843CbZsRfzWp5FTusNXXL1o4N3jd0hoCNpsoUR6Qjxk,23306
52
52
  sglang/srt/layers/attention/triton_backend.py,sha256=DKUEzxQE8iBvJPNHmQwP1pyx2wXmSsLqzBhLjJznIUk,6482
53
53
  sglang/srt/layers/attention/triton_ops/decode_attention.py,sha256=XCQTX0kUttT1AG5FRMgfQbiXgvoempYD0UR2r6D_vJg,16711
54
54
  sglang/srt/layers/attention/triton_ops/double_sparsity_attention.py,sha256=1pSXfY3EEaM7iRN_uElHnAfsrJMhTFbu9fj8Z0O2PbE,21480
@@ -63,24 +63,24 @@ sglang/srt/layers/quantization/base_config.py,sha256=vlpSPvSrFmUe65ETg4SoPocQ9bV
63
63
  sglang/srt/lora/lora.py,sha256=meRL7oBUx8mxV_isc3Lp0EIsFQWC2PvaN-fE78BmMwg,14970
64
64
  sglang/srt/lora/lora_config.py,sha256=paVB7F7SIuxr_vodvKf8zzAlH2fdVYHhXxcXV62D0Vo,1411
65
65
  sglang/srt/lora/lora_manager.py,sha256=gzBwYXZEPYj56PkGTshTbWRfl_370wb6uTcRhDaLiF8,12801
66
- sglang/srt/managers/data_parallel_controller.py,sha256=GJGfX1-5DoQFZ-EMh_p02nvrOtrOc0UebnULWHhFrss,5765
67
- sglang/srt/managers/detokenizer_manager.py,sha256=G2-Y-nDbq7LF8ZFWcXXcMkbCwzyBEh1g4UrciDlkNYY,7985
66
+ sglang/srt/managers/data_parallel_controller.py,sha256=1MqXT5t9NHsnrTj-b_SUfFNYgKk-6lUdbqpqS1EbftM,5766
67
+ sglang/srt/managers/detokenizer_manager.py,sha256=wgFtHwmx336tUNa-BWBmdf7t3qSj0zHpdHI4pmTiqbY,8163
68
68
  sglang/srt/managers/image_processor.py,sha256=yhrXZvqPq5iSjycez7pA7YyzDS1VAXvCl5IuQ8uKwfs,13793
69
- sglang/srt/managers/io_struct.py,sha256=QJ1Eu7XNsRH35ec5nUOUS2XSjiFWyjYHKsnMqviC_Mk,12298
70
- sglang/srt/managers/schedule_batch.py,sha256=Fo-SbHZaasW8ec4EGFZV4BT447rCxOFkd6wact0JFfQ,39913
69
+ sglang/srt/managers/io_struct.py,sha256=bEvbibB2xJLFLUxgjCl5VgOBT1VtaH5VAv3zYlirxKY,12403
70
+ sglang/srt/managers/schedule_batch.py,sha256=0kZnX1ciaL9yXxUPRyGsEArDvgHnTrZ2V8X4So7VJws,40509
71
71
  sglang/srt/managers/schedule_policy.py,sha256=unDmK7Y_Ti0Eiizh3_iEFMsC1KDqGMTqU8MlQgg-6qo,11951
72
- sglang/srt/managers/scheduler.py,sha256=ETFjzp5mMqeRrICzn44C-SksWC96hTGDM9WQpdf_QJU,43523
73
- sglang/srt/managers/tokenizer_manager.py,sha256=9culfJz4O1m18pye0S-hexrP86GVtJJyPjMXPt_-qGc,25502
72
+ sglang/srt/managers/scheduler.py,sha256=e1lYL3Ky8c4Ei6h3lG-r2SavyZlymFMBEn37eFvjsso,43821
73
+ sglang/srt/managers/tokenizer_manager.py,sha256=6KaDIzv3yiyRmLOVzvol7rTgnOINKgynLUrRHEXm7Xo,25980
74
74
  sglang/srt/managers/tp_worker.py,sha256=LbVXE6GiCnmDr117HjB-lxVbFqLL71YEM9y0Q46nWns,5733
75
- sglang/srt/managers/tp_worker_overlap_thread.py,sha256=zxTiXlhI9KWDjVSdM8enxYZKLe-GQlCQjfNSMxcZ2uY,5950
75
+ sglang/srt/managers/tp_worker_overlap_thread.py,sha256=j5J4yHyR7w2HgAbN7S__299ADvsoyap5HK63SWMNavQ,7546
76
76
  sglang/srt/mem_cache/base_prefix_cache.py,sha256=qEQwEkG4E5rab2ZoTqcesf5pR_J4nV2jBxIHsBJHtIM,924
77
77
  sglang/srt/mem_cache/chunk_cache.py,sha256=VcCpyrf5FOQ5xoKeOouCI5ZQLkZo_pgY1SPbDDkagGg,2492
78
78
  sglang/srt/mem_cache/flush_cache.py,sha256=pTLKPRB17U6vl5RFJJvuJ4jCL2SyomgkUBNlkDpGRqo,978
79
- sglang/srt/mem_cache/memory_pool.py,sha256=bMPA0M_yIiIPooFp-xQDQV7mHhA957iaIv7eVYx3h9k,10606
79
+ sglang/srt/mem_cache/memory_pool.py,sha256=41fjuj_sD0yfJq-sy-X99cc2djBa6w4dy2y47V0WqNU,10934
80
80
  sglang/srt/mem_cache/radix_cache.py,sha256=DzLCO_gYQ7X_C2NJSEHzzMZhb5HzWjKF9wXJQsnzr8M,10427
81
- sglang/srt/model_executor/cuda_graph_runner.py,sha256=Upuz2ftg4rVaCwQ3i3T9g40KTkwK3f8XbYTGxcn_AHg,12472
82
- sglang/srt/model_executor/forward_batch_info.py,sha256=ilDs3UaZ6_euAAQs8x0vg2WKxBnAgd21fXsspaV6_68,9442
83
- sglang/srt/model_executor/model_runner.py,sha256=7Jd4Ukt_r20rtMHuV5D_rdZnqhNoLLQX54xRydZ7RfU,27030
81
+ sglang/srt/model_executor/cuda_graph_runner.py,sha256=t2KPOtoz1k-LhEs8QOHNFgz-tMslI0TSlyBX74WRuaU,12828
82
+ sglang/srt/model_executor/forward_batch_info.py,sha256=1GM6A-tqTDD0MEMQx93PC7XahABr0vlv7JBXohaehkc,9272
83
+ sglang/srt/model_executor/model_runner.py,sha256=90eHOkVSszaXwBgTowCl4pTrzZyJAIosnb2E6MuUXd8,27038
84
84
  sglang/srt/models/baichuan.py,sha256=uV20fr7SqlATxoziXRtJyXFnZZoWTUy3qvQNFaEvw8M,15014
85
85
  sglang/srt/models/chatglm.py,sha256=lIu_pVtIWNJ8EsZfr2KZdr2m4GtDTUNvz6D4zkpLMq4,13144
86
86
  sglang/srt/models/commandr.py,sha256=WIMwjV3C0pRbVs4Xv9tqnHGreRvWC7zsML2hNuXw4A0,14060
@@ -119,16 +119,16 @@ sglang/srt/models/yivl.py,sha256=xcWqkuZ29FmBBJY6aKetwItWIPl-kfXK-QmgdLONles,476
119
119
  sglang/srt/openai_api/adapter.py,sha256=WkYCKVaYTkFdLrySBhlkDyHJVaaHMF7KrhNnmw3L3us,53534
120
120
  sglang/srt/openai_api/protocol.py,sha256=EZ6G209rBEDP7cepO2kAYqE8wMe1ksYdN7to1iT97Lw,10248
121
121
  sglang/srt/sampling/sampling_batch_info.py,sha256=dT81cFCe8zKor7P3g4ITnELTySX2nKX_FoNIU8y78Zo,7911
122
- sglang/srt/sampling/sampling_params.py,sha256=ZPHCQq7Bi4P_sxUzdKgYVXZpB_tC-kA7rlLwiW9Ct9A,5781
122
+ sglang/srt/sampling/sampling_params.py,sha256=u1UWt9biIFXKymAg56RbkMa8oe5jxsoMvsv3cH7_kZ0,5692
123
123
  sglang/srt/sampling/penaltylib/__init__.py,sha256=5vQw0Y5DSzmsoFg1IdMIKLwFVhYZ5ArADHVBYbSmOec,513
124
124
  sglang/srt/sampling/penaltylib/orchestrator.py,sha256=kizcPnxtRawmDt6utRuhbk4yfNs5H5mx1DAlDVEZRv8,11328
125
125
  sglang/srt/sampling/penaltylib/penalizers/frequency_penalty.py,sha256=IvYioX53Vq_ji-0Zhcz_r5mUa3T3GaIydVS6K4FhWfE,2557
126
- sglang/srt/sampling/penaltylib/penalizers/min_new_tokens.py,sha256=XJZP0C4NFyXgcODbIWXxrgVEjmRgqLdZuVAtoN-LveY,3565
126
+ sglang/srt/sampling/penaltylib/penalizers/min_new_tokens.py,sha256=MmfqRqJ-leSoY9iO5Hg_ILlX-M0M0tObYrxrb_quStg,3717
127
127
  sglang/srt/sampling/penaltylib/penalizers/presence_penalty.py,sha256=0PlANTrR959foTA3Nj5qBE7ndaOZgG-9X6LhzlmEUc8,2533
128
128
  sglang/srt/sampling/penaltylib/penalizers/repetition_penalty.py,sha256=v9jOgA0-I31WcrhIydiFbpy2ZJPLytFLGM98NRPd2sU,2820
129
129
  sglang/test/few_shot_gsm8k.py,sha256=ll-gNbcv829IwSPXAZt4JIEIu8IR3APCLcX3BHOFVp8,3968
130
130
  sglang/test/few_shot_gsm8k_engine.py,sha256=QQbrwOX6-cJDD3RZC_e7zPnt6aSo8JdF8X_lRHSjdDM,3886
131
- sglang/test/run_eval.py,sha256=NWxeLWmInBgkCvC9Jr_QzF7GfAiBve3Gf1JQrEOlNlU,3899
131
+ sglang/test/run_eval.py,sha256=9yO0hXZOcn4abEOs96T-XPguDEklK16Ltco0pGF3zCg,4020
132
132
  sglang/test/runners.py,sha256=zhv2CGKnJjhXNK6Y0U8uV_4g30QGtmO2KKxUohVgibw,14246
133
133
  sglang/test/simple_eval_common.py,sha256=r0G-9QLycs2ax3RMc44T_61fzMxlpTzv6pececC7lyY,12379
134
134
  sglang/test/simple_eval_gpqa.py,sha256=8Xt9Bw05c7SZTYrCZgB68OZUqUbLo69ywiyx0bTvSUk,3220
@@ -139,10 +139,10 @@ sglang/test/simple_eval_mmlu.py,sha256=FkwamjGMjueTixymkedF-YiPloSLiy4ftILFUrKZ9
139
139
  sglang/test/test_activation.py,sha256=jkdNRzJnbd5OgZliQaIXpxovlcky17UrweomcOcMxoE,1442
140
140
  sglang/test/test_layernorm.py,sha256=IacByD5d-stXjzBz8Ypamc7povlcedpKPbb_4JLgo3c,3720
141
141
  sglang/test/test_programs.py,sha256=1Z0umrsUu9pagzyGH5SrXl_qhKSyTfUv_kWC2mcn0qo,18208
142
- sglang/test/test_utils.py,sha256=NkJuezjmonjgC3_i_CTBd8KSqWh6W9CLcgoaqvTNK2U,18684
143
- sglang/test/srt/sampling/penaltylib/utils.py,sha256=Koe8GYoxIBUCz71of0oHhM5t5QcEd6a1IYq5SszRFAw,12730
144
- sglang-0.3.4.post1.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
145
- sglang-0.3.4.post1.dist-info/METADATA,sha256=Ztpjtx7hLwak_B8teVM_z1tANGZzyraV_5hcHTx3hg4,40804
146
- sglang-0.3.4.post1.dist-info/WHEEL,sha256=OVMc5UfuAQiSplgO0_WdW7vXVGAt9Hdd6qtN4HotdyA,91
147
- sglang-0.3.4.post1.dist-info/top_level.txt,sha256=yxhh3pYQkcnA7v3Bg889C2jZhvtJdEincysO7PEB09M,7
148
- sglang-0.3.4.post1.dist-info/RECORD,,
142
+ sglang/test/test_utils.py,sha256=NXbw0qyv0OpWMHAT0g_RDYiEYcjhTf1Ud5m-gRrX4kY,21234
143
+ sglang/test/srt/sampling/penaltylib/utils.py,sha256=q98pQDikkmvvvvAG-AXMYaYte1iHHW2TFhKGtAeGvdE,12802
144
+ sglang-0.3.4.post2.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
145
+ sglang-0.3.4.post2.dist-info/METADATA,sha256=wC1SzkNFr9fGqhzFDqn6yRrjhIBgXtxsaEDYgekz0RQ,41431
146
+ sglang-0.3.4.post2.dist-info/WHEEL,sha256=OVMc5UfuAQiSplgO0_WdW7vXVGAt9Hdd6qtN4HotdyA,91
147
+ sglang-0.3.4.post2.dist-info/top_level.txt,sha256=yxhh3pYQkcnA7v3Bg889C2jZhvtJdEincysO7PEB09M,7
148
+ sglang-0.3.4.post2.dist-info/RECORD,,