sglang 0.4.4__py3-none-any.whl → 0.4.4.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 (176) hide show
  1. sglang/__init__.py +2 -0
  2. sglang/api.py +6 -0
  3. sglang/bench_one_batch.py +1 -1
  4. sglang/bench_one_batch_server.py +1 -1
  5. sglang/bench_serving.py +3 -1
  6. sglang/check_env.py +3 -4
  7. sglang/lang/backend/openai.py +18 -5
  8. sglang/lang/chat_template.py +28 -7
  9. sglang/lang/interpreter.py +7 -3
  10. sglang/lang/ir.py +10 -0
  11. sglang/srt/_custom_ops.py +1 -1
  12. sglang/srt/code_completion_parser.py +174 -0
  13. sglang/srt/configs/__init__.py +2 -6
  14. sglang/srt/configs/deepseekvl2.py +667 -0
  15. sglang/srt/configs/janus_pro.py +3 -4
  16. sglang/srt/configs/load_config.py +1 -0
  17. sglang/srt/configs/model_config.py +63 -11
  18. sglang/srt/configs/utils.py +25 -0
  19. sglang/srt/connector/__init__.py +51 -0
  20. sglang/srt/connector/base_connector.py +112 -0
  21. sglang/srt/connector/redis.py +85 -0
  22. sglang/srt/connector/s3.py +122 -0
  23. sglang/srt/connector/serde/__init__.py +31 -0
  24. sglang/srt/connector/serde/safe_serde.py +29 -0
  25. sglang/srt/connector/serde/serde.py +43 -0
  26. sglang/srt/connector/utils.py +35 -0
  27. sglang/srt/conversation.py +88 -0
  28. sglang/srt/disaggregation/conn.py +81 -0
  29. sglang/srt/disaggregation/decode.py +495 -0
  30. sglang/srt/disaggregation/mini_lb.py +285 -0
  31. sglang/srt/disaggregation/prefill.py +249 -0
  32. sglang/srt/disaggregation/utils.py +44 -0
  33. sglang/srt/distributed/parallel_state.py +10 -3
  34. sglang/srt/entrypoints/engine.py +55 -5
  35. sglang/srt/entrypoints/http_server.py +71 -12
  36. sglang/srt/function_call_parser.py +164 -54
  37. sglang/srt/hf_transformers_utils.py +28 -3
  38. sglang/srt/layers/activation.py +4 -2
  39. sglang/srt/layers/attention/base_attn_backend.py +1 -1
  40. sglang/srt/layers/attention/flashattention_backend.py +295 -0
  41. sglang/srt/layers/attention/flashinfer_backend.py +1 -1
  42. sglang/srt/layers/attention/flashmla_backend.py +284 -0
  43. sglang/srt/layers/attention/triton_backend.py +171 -38
  44. sglang/srt/layers/attention/triton_ops/decode_attention.py +94 -31
  45. sglang/srt/layers/attention/triton_ops/extend_attention.py +14 -5
  46. sglang/srt/layers/attention/utils.py +53 -0
  47. sglang/srt/layers/attention/vision.py +9 -28
  48. sglang/srt/layers/dp_attention.py +62 -23
  49. sglang/srt/layers/elementwise.py +411 -0
  50. sglang/srt/layers/layernorm.py +24 -2
  51. sglang/srt/layers/linear.py +17 -5
  52. sglang/srt/layers/logits_processor.py +26 -7
  53. sglang/srt/layers/moe/ep_moe/kernels.py +110 -11
  54. sglang/srt/layers/moe/ep_moe/layer.py +273 -1
  55. sglang/srt/layers/moe/ep_moe/token_dispatcher.py +416 -0
  56. sglang/srt/layers/moe/fused_moe_native.py +2 -1
  57. sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=64,device_name=NVIDIA_L20,dtype=int8_w8a8.json +146 -0
  58. sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=64,device_name=NVIDIA_L40S,dtype=int8_w8a8.json +146 -0
  59. sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=1024,device_name=NVIDIA_H100_80GB_HBM3,dtype=fp8_w8a8,block_shape=[128, 128].json +146 -0
  60. sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=512,device_name=NVIDIA_H200,dtype=fp8_w8a8,block_shape=[128, 128].json +146 -0
  61. sglang/srt/layers/moe/fused_moe_triton/fused_moe.py +23 -32
  62. sglang/srt/layers/moe/fused_moe_triton/layer.py +1 -2
  63. sglang/srt/layers/moe/router.py +342 -0
  64. sglang/srt/layers/moe/topk.py +31 -18
  65. sglang/srt/layers/parameter.py +1 -1
  66. sglang/srt/layers/quantization/__init__.py +184 -126
  67. sglang/srt/layers/quantization/base_config.py +5 -0
  68. sglang/srt/layers/quantization/blockwise_int8.py +1 -1
  69. sglang/srt/layers/quantization/compressed_tensors/__init__.py +0 -0
  70. sglang/srt/layers/quantization/compressed_tensors/compressed_tensors.py +652 -0
  71. sglang/srt/layers/quantization/compressed_tensors/compressed_tensors_moe.py +658 -0
  72. sglang/srt/layers/quantization/compressed_tensors/schemes/__init__.py +9 -0
  73. sglang/srt/layers/quantization/compressed_tensors/schemes/compressed_tensors_scheme.py +56 -0
  74. sglang/srt/layers/quantization/compressed_tensors/schemes/compressed_tensors_w8a8_fp8.py +162 -0
  75. sglang/srt/layers/quantization/compressed_tensors/utils.py +218 -0
  76. sglang/srt/layers/quantization/fp8.py +76 -34
  77. sglang/srt/layers/quantization/fp8_kernel.py +24 -8
  78. sglang/srt/layers/quantization/fp8_utils.py +284 -28
  79. sglang/srt/layers/quantization/gptq.py +36 -9
  80. sglang/srt/layers/quantization/kv_cache.py +98 -0
  81. sglang/srt/layers/quantization/modelopt_quant.py +9 -7
  82. sglang/srt/layers/quantization/utils.py +153 -0
  83. sglang/srt/layers/quantization/w8a8_fp8.py +70 -19
  84. sglang/srt/layers/rotary_embedding.py +66 -87
  85. sglang/srt/layers/sampler.py +1 -1
  86. sglang/srt/lora/layers.py +68 -0
  87. sglang/srt/lora/lora.py +2 -22
  88. sglang/srt/lora/lora_manager.py +47 -23
  89. sglang/srt/lora/mem_pool.py +110 -51
  90. sglang/srt/lora/utils.py +12 -1
  91. sglang/srt/managers/cache_controller.py +4 -5
  92. sglang/srt/managers/data_parallel_controller.py +31 -9
  93. sglang/srt/managers/expert_distribution.py +81 -0
  94. sglang/srt/managers/io_struct.py +39 -3
  95. sglang/srt/managers/mm_utils.py +373 -0
  96. sglang/srt/managers/multimodal_processor.py +68 -0
  97. sglang/srt/managers/multimodal_processors/base_processor.py +275 -0
  98. sglang/srt/managers/multimodal_processors/deepseek_vl_v2.py +119 -0
  99. sglang/srt/managers/multimodal_processors/gemma3.py +83 -0
  100. sglang/srt/managers/{image_processors → multimodal_processors}/janus_pro.py +20 -15
  101. sglang/srt/managers/{image_processors → multimodal_processors}/llava.py +10 -15
  102. sglang/srt/managers/multimodal_processors/minicpm.py +167 -0
  103. sglang/srt/managers/{image_processors → multimodal_processors}/mlama.py +7 -8
  104. sglang/srt/managers/{image_processors → multimodal_processors}/qwen_vl.py +28 -22
  105. sglang/srt/managers/schedule_batch.py +134 -31
  106. sglang/srt/managers/scheduler.py +325 -38
  107. sglang/srt/managers/scheduler_output_processor_mixin.py +4 -1
  108. sglang/srt/managers/session_controller.py +1 -1
  109. sglang/srt/managers/tokenizer_manager.py +59 -23
  110. sglang/srt/managers/tp_worker.py +1 -1
  111. sglang/srt/managers/tp_worker_overlap_thread.py +3 -3
  112. sglang/srt/managers/utils.py +6 -1
  113. sglang/srt/mem_cache/hiradix_cache.py +27 -8
  114. sglang/srt/mem_cache/memory_pool.py +258 -98
  115. sglang/srt/mem_cache/paged_allocator.py +2 -2
  116. sglang/srt/mem_cache/radix_cache.py +4 -4
  117. sglang/srt/model_executor/cuda_graph_runner.py +85 -28
  118. sglang/srt/model_executor/forward_batch_info.py +81 -15
  119. sglang/srt/model_executor/model_runner.py +70 -6
  120. sglang/srt/model_loader/loader.py +160 -2
  121. sglang/srt/model_loader/weight_utils.py +45 -0
  122. sglang/srt/models/deepseek_janus_pro.py +29 -86
  123. sglang/srt/models/deepseek_nextn.py +22 -10
  124. sglang/srt/models/deepseek_v2.py +326 -192
  125. sglang/srt/models/deepseek_vl2.py +358 -0
  126. sglang/srt/models/gemma3_causal.py +684 -0
  127. sglang/srt/models/gemma3_mm.py +462 -0
  128. sglang/srt/models/grok.py +374 -119
  129. sglang/srt/models/llama.py +47 -7
  130. sglang/srt/models/llama_eagle.py +1 -0
  131. sglang/srt/models/llama_eagle3.py +196 -0
  132. sglang/srt/models/llava.py +3 -3
  133. sglang/srt/models/llavavid.py +3 -3
  134. sglang/srt/models/minicpmo.py +1995 -0
  135. sglang/srt/models/minicpmv.py +62 -137
  136. sglang/srt/models/mllama.py +4 -4
  137. sglang/srt/models/phi3_small.py +1 -1
  138. sglang/srt/models/qwen2.py +3 -0
  139. sglang/srt/models/qwen2_5_vl.py +68 -146
  140. sglang/srt/models/qwen2_classification.py +75 -0
  141. sglang/srt/models/qwen2_moe.py +9 -1
  142. sglang/srt/models/qwen2_vl.py +25 -63
  143. sglang/srt/openai_api/adapter.py +145 -47
  144. sglang/srt/openai_api/protocol.py +23 -2
  145. sglang/srt/sampling/sampling_batch_info.py +1 -1
  146. sglang/srt/sampling/sampling_params.py +6 -6
  147. sglang/srt/server_args.py +104 -14
  148. sglang/srt/speculative/build_eagle_tree.py +7 -347
  149. sglang/srt/speculative/eagle_draft_cuda_graph_runner.py +41 -5
  150. sglang/srt/speculative/eagle_utils.py +208 -252
  151. sglang/srt/speculative/eagle_worker.py +139 -53
  152. sglang/srt/speculative/spec_info.py +6 -1
  153. sglang/srt/torch_memory_saver_adapter.py +22 -0
  154. sglang/srt/utils.py +182 -21
  155. sglang/test/__init__.py +0 -0
  156. sglang/test/attention/__init__.py +0 -0
  157. sglang/test/attention/test_flashattn_backend.py +312 -0
  158. sglang/test/runners.py +2 -0
  159. sglang/test/test_activation.py +2 -1
  160. sglang/test/test_block_fp8.py +5 -4
  161. sglang/test/test_block_fp8_ep.py +2 -1
  162. sglang/test/test_dynamic_grad_mode.py +58 -0
  163. sglang/test/test_layernorm.py +3 -2
  164. sglang/test/test_utils.py +55 -4
  165. sglang/utils.py +31 -0
  166. sglang/version.py +1 -1
  167. {sglang-0.4.4.dist-info → sglang-0.4.4.post2.dist-info}/METADATA +12 -8
  168. {sglang-0.4.4.dist-info → sglang-0.4.4.post2.dist-info}/RECORD +171 -125
  169. {sglang-0.4.4.dist-info → sglang-0.4.4.post2.dist-info}/WHEEL +1 -1
  170. sglang/srt/configs/qwen2_5_vl_config.py +0 -1006
  171. sglang/srt/managers/image_processor.py +0 -55
  172. sglang/srt/managers/image_processors/base_image_processor.py +0 -219
  173. sglang/srt/managers/image_processors/minicpmv.py +0 -86
  174. sglang/srt/managers/multi_modality_padding.py +0 -134
  175. {sglang-0.4.4.dist-info → sglang-0.4.4.post2.dist-info/licenses}/LICENSE +0 -0
  176. {sglang-0.4.4.dist-info → sglang-0.4.4.post2.dist-info}/top_level.txt +0 -0
@@ -1,60 +1,75 @@
1
- sglang/__init__.py,sha256=njc4c2IBYklSqVMiT70GL630Uddg5D_IU_6dthApPxc,1587
2
- sglang/api.py,sha256=rYa2qKE88_RJQwYVvjuJzEZECf75ujchZVqi0q48tqc,6890
1
+ sglang/__init__.py,sha256=8J5PHcfRJul4R1NJnt0BtW05aVQNWqp8PyalLPOTZCA,1669
2
+ sglang/api.py,sha256=vHiKBg8wwIdmrpnGclop5BzJ-1Q88emrlrfLwNCHg98,7010
3
3
  sglang/bench_offline_throughput.py,sha256=OQb-AjL4UNymmir02ht43uzgaNsnO_I11nXSowKMqBI,13841
4
- sglang/bench_one_batch.py,sha256=mVgmg1LP_Y67HlH4M2V7QvuO1aFjCE0n3gDRXW_w_NQ,17935
5
- sglang/bench_one_batch_server.py,sha256=iu73SsvYwnuRktYZDz1P6psMiRx8MbEbF5sbsYJdzYg,5962
6
- sglang/bench_serving.py,sha256=ScnPmhg4S5qYJZ0GdwJZf4Eb-j3EVS-gMRQAnFoQqIs,55355
7
- sglang/check_env.py,sha256=lDVA3ybt1wOE33HIMpkkU7zGRgLWez1_ifRRJ8qxbtw,8445
4
+ sglang/bench_one_batch.py,sha256=Fp6HBBJHrw672Q1gnklJ7dYboYYjR92D2fNCvbrM3M0,17935
5
+ sglang/bench_one_batch_server.py,sha256=8VYNhaQbWGP8TkNVuy_sPjD5FiuVZHamtGRWKwa-Z-Q,5962
6
+ sglang/bench_serving.py,sha256=VR04tzBC3RnKrEsDp6Vm5L_t_lVrsV1ERA_GESPVFbw,55398
7
+ sglang/check_env.py,sha256=76itNLUw9KlqbiY1BI4u4YaMZaqyCNcrCLUIb6aHflM,8396
8
8
  sglang/global_config.py,sha256=xzLdk8W53fneFblNh8iIjGF9C3-7mnzR1-LleD9Btxg,1495
9
9
  sglang/launch_server.py,sha256=mDXfwha8LHpWQJekcCosR98QhCQsbmilsBlI5jAIgg0,420
10
10
  sglang/llama3_eval.py,sha256=gWSboDchIGybIce88bJlrCG0yiLZ513mw4gcutJlzGM,10017
11
- sglang/utils.py,sha256=jMMRaAChfWgU_F2BMbDF9G896shiemdzCnWxFocy4QM,15195
12
- sglang/version.py,sha256=6G_giX6Ucuweo7w5OiftoXmbNLoqiU_soXJoU8aiLmY,22
11
+ sglang/utils.py,sha256=Ns1-swsHRb8jYA0wWCQjCR_CDBWIJgyoooEng4fsqQc,16155
12
+ sglang/version.py,sha256=v-O7_0X8kFcUc6ds3KpBMpI4Tafk-LT7SJxJ_Q2a3MU,28
13
13
  sglang/lang/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
14
- sglang/lang/chat_template.py,sha256=VKI5RNydfo1-qz6HT-aYm6ecBsa1RjkTZLZLCiBb5Ig,18269
14
+ sglang/lang/chat_template.py,sha256=xZ-kQpgb4-NY6QOqgRjOODZW5G4EvJnplaqYF6Ng2Ow,18952
15
15
  sglang/lang/choices.py,sha256=-W1DVw9N9ZliVpvmWrzIXG4cswAah8eMQrHWzkS3D8o,6234
16
16
  sglang/lang/compiler.py,sha256=o1C6G3TzhjSlsH-doTPy5oiVehr57dxNTa5oZw5TTAI,7639
17
- sglang/lang/interpreter.py,sha256=r7x5mBxAOaEwmxjaMBMcn7N8HDFv6V6K9eINtffDygQ,33074
18
- sglang/lang/ir.py,sha256=fUQc7ExghlQNGFxYpD-uLXonVSCA3GUGdd8PNLPIP5Y,18560
17
+ sglang/lang/interpreter.py,sha256=OH1SFCm4rUCPO32MTo8j5V2Z13Jic7_r1GQOP1-aHaw,33234
18
+ sglang/lang/ir.py,sha256=gssBGxqQEVJmjR_PqG2yah48AsInGaO3CmnQ2boJThc,18769
19
19
  sglang/lang/tracer.py,sha256=o-jLAPPSuy2vBfsGGrTAnbuWtORzQ50B4C_P5zvYkx8,8291
20
20
  sglang/lang/backend/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
21
21
  sglang/lang/backend/anthropic.py,sha256=EXRX7xJgA5KZszX7toSLVnKzFQ5EO0Loj-YjHFtxSxg,2081
22
22
  sglang/lang/backend/base_backend.py,sha256=tdoh9YF3CyekY1BKiX9n7-aA4srDWIuA4RDJLM7q8qg,1985
23
23
  sglang/lang/backend/litellm.py,sha256=ugmL7sfUxkUHVbHtwNzHgdQAEd4UCjNQboFuE3KThcY,2450
24
- sglang/lang/backend/openai.py,sha256=BQj1FHPXmSfFVQV-SIs7WW6v7tUDUckjtpvs9mhP8Ok,15645
24
+ sglang/lang/backend/openai.py,sha256=6I1udBC9obVpJG0GerbFVbWCbwLKkF2DoPdHBCnqPzs,16341
25
25
  sglang/lang/backend/runtime_endpoint.py,sha256=CAVh3X9F80t_2tkJECF__7AdCQtqDg1AHDqIoKIPnvs,16755
26
26
  sglang/lang/backend/vertexai.py,sha256=O-iBLD-y3vq80UxnrAoJri7bxpgd-_eakZ88Cf8bEGA,4855
27
- sglang/srt/_custom_ops.py,sha256=0zKNhmLhtUNTBE_WwuzkG_Y_Kaufai84VGe4fJ8v6-o,4975
27
+ sglang/srt/_custom_ops.py,sha256=QumGYYWh5qiwNdg9Uj263Rq26oFX7ldB2hcaqDEQN_k,4976
28
28
  sglang/srt/aio_rwlock.py,sha256=6LYtOdeTUY3hkfa1dmYkgsaF2ttrwIF3hUWz2AZ2fqw,2970
29
- sglang/srt/conversation.py,sha256=qiOAs42rftnIgjPq5c8Ce0DSCMsaRlvFxyUTpUXR_4M,24664
29
+ sglang/srt/code_completion_parser.py,sha256=HhEUzdL-FVBsOot9tKDKA1l8Gdx8qsF1RRg-zHNpmLQ,5400
30
+ sglang/srt/conversation.py,sha256=mzNPZX7ndgmm1E8azuK8eytN6bNCEu3WbcPReRFfhd0,27819
30
31
  sglang/srt/custom_op.py,sha256=vhdofFbWtpdtaA4NG4tkanWwEsfvBnjh6OPKOxmxXdU,3426
31
- sglang/srt/function_call_parser.py,sha256=-siKVUqr3B3pufJ8G0wKGoQKQniAcCa8K7x1kFgQyyQ,19532
32
- sglang/srt/hf_transformers_utils.py,sha256=FPoOvGdEy-lwrOOqHaQnb_tJCGmCW3CpKUjleY7Cwj4,8221
32
+ sglang/srt/function_call_parser.py,sha256=8fWRVeOZ60IjpCwjbEvpCID7tWR_ls7kVvDjvaJe0wQ,23778
33
+ sglang/srt/hf_transformers_utils.py,sha256=_QYTl9LpU0jmKPlYooHi1etwMvb5v40JIrG_t_Fx06w,9215
33
34
  sglang/srt/mm_utils.py,sha256=1ScBunw_x4W8ebM_AcJ62-1T2mfT8NlMJqdAhkF1lb0,12367
34
35
  sglang/srt/model_parallel.py,sha256=eLXZhvJ4wG6dh0FontNCIdVZvHYdWgaeY-5cu7TD9tE,6078
35
36
  sglang/srt/reasoning_parser.py,sha256=45xsU9RCPfyG4_Zx4y3-JPyNgAtrqwKI4j5R2NT4g1s,5594
36
37
  sglang/srt/server.py,sha256=PrQb9r6L9syWHKlggbbiQYsKtpwSmECqozRbf8qnoV8,874
37
- sglang/srt/server_args.py,sha256=GR_SuD3tttfwCJoRi8OHFICsTqLN0CexOIBpXZoxrww,44949
38
- sglang/srt/torch_memory_saver_adapter.py,sha256=--FgbrcvJxTcRe856plD9ktqgrHGPTE18eZCJlE50hY,1255
39
- sglang/srt/utils.py,sha256=Nh85Igdbmi5nN09s2rFsnnAIUcijk5jljmktiYByZp8,50163
38
+ sglang/srt/server_args.py,sha256=40PhLjwRoBqS0q_lKPZ0YGxtfw5BU3BpuyosmJ7WgMQ,48617
39
+ sglang/srt/torch_memory_saver_adapter.py,sha256=KG3wM9-xZsSdsmORofArnNR7hH55GEyFxaderCDcK9w,1853
40
+ sglang/srt/utils.py,sha256=19-F9YV5dYqf_kVJ-OiVNO_xWTZBNUv-uQSxhTXMEZg,55093
40
41
  sglang/srt/warmup.py,sha256=FmJiYfjRr3X_eAe7ojQaPoN17LvHpjDmRWRnO-k86AQ,1469
41
- sglang/srt/configs/__init__.py,sha256=iDfWab2iDMQ3F5bezHjM-fHi7u8iKqk_UBLZFabXxiI,470
42
+ sglang/srt/configs/__init__.py,sha256=vulncVn70WqIT6s0HaB8p_Q6FjOiaLwNZWpoJS9FIuQ,399
42
43
  sglang/srt/configs/chatglm.py,sha256=j-b0YkdYUmQm2y1kNmMJtKeACxWKmBbvNNkDWbs6kbI,2907
43
44
  sglang/srt/configs/dbrx.py,sha256=tdhIkXAQl1yr0MxqFmsDG1E0e2puRTTKm6UTyANBLac,11005
45
+ sglang/srt/configs/deepseekvl2.py,sha256=2CcUemLusSoqdaZ6bzQiSCxVzbB3O_o9e3hsWgaTC_Q,23013
44
46
  sglang/srt/configs/device_config.py,sha256=kfmpPOECqYxcRoY-ko0QZRhyiBWUGP2CMF51DMUN5nU,435
45
47
  sglang/srt/configs/exaone.py,sha256=Duxd4yQoKy8GWEzZD_kCY_OzmN_67CTJL_Kgn0eXk3g,10731
46
- sglang/srt/configs/janus_pro.py,sha256=v034OiY264vMmGvanHuNL6FzikgqTVwyUoKO26c-PQ0,19015
47
- sglang/srt/configs/load_config.py,sha256=NcptQBTG-TBJpfcnE1e32BefBBzrINm9n5bYzTrH1bs,3287
48
- sglang/srt/configs/model_config.py,sha256=3v43xDudZkF1AEOZzYHNsq3FDsEBW2-MWSduggDYOE4,18748
49
- sglang/srt/configs/qwen2_5_vl_config.py,sha256=x7erMMDwqlI2l6YYBvZCXlZ53li2waWWgyhJjz273dk,48223
48
+ sglang/srt/configs/janus_pro.py,sha256=-QtJ4ZGZiAJb0AkOEcuCHzIKLw23nF8nRk3rdCcoUO0,19016
49
+ sglang/srt/configs/load_config.py,sha256=qs-AxuplouBx2tsv9KGBOLZPbwzuVA4vbktbGP_cRp8,3309
50
+ sglang/srt/configs/model_config.py,sha256=M2iEdfn2g_NkAHwmUHMPhGbXzWoNz1H_pPNcVXZFmQs,20947
51
+ sglang/srt/configs/utils.py,sha256=3nHUfisMs_Ltuhv8OZTNCJp63YJKJVF43h1QZB1zqx8,670
52
+ sglang/srt/connector/__init__.py,sha256=czLX5JOxuMhH-T9eSJzoc1qv1B4z9chyffDRL5I6wo4,1247
53
+ sglang/srt/connector/base_connector.py,sha256=i6i1TIzsz4NbSEkrdMPq-urb2sN2aLAx8dazga4gB9U,2833
54
+ sglang/srt/connector/redis.py,sha256=KoLLUSY4Lu9zUrRJy2yYdtf1yanTaS0bQP8V0t95DB0,2552
55
+ sglang/srt/connector/s3.py,sha256=hn5Ow9Y6ItS8RMJDwnTQe5QaiQ7HKMgiwggE9QnUewM,3743
56
+ sglang/srt/connector/utils.py,sha256=isTvxauz1_8no5MW6p8Bwj2H9mQkweaRO_RSwAPA7R8,983
57
+ sglang/srt/connector/serde/__init__.py,sha256=MvBJ7FBJtnou-AD-EdsCdAIDCcA8kWvUNuRViILVJ-0,718
58
+ sglang/srt/connector/serde/safe_serde.py,sha256=Iv_mFsDvk-cXuw0WTykBZ2yGxW0jb82DwVqC7r3RmAU,750
59
+ sglang/srt/connector/serde/serde.py,sha256=n59I2MXLa7WCyN_8pEd8L-scJk7lMhmEX-GOUIhF0ZA,1004
50
60
  sglang/srt/constrained/base_grammar_backend.py,sha256=q2RTH9hv2nKUF_-HVdZzEUjG6LxejPqXjvUOE1NDWaU,6788
51
61
  sglang/srt/constrained/llguidance_backend.py,sha256=ej7wN13SzCsT310C6OIyUg2zs5jeuLl3Ocok9SP9-c4,5702
52
62
  sglang/srt/constrained/outlines_backend.py,sha256=UWv2xjg8x4XtoqpY8LoorlJaYOZhfDeIr5YCiFn4knA,6812
53
63
  sglang/srt/constrained/outlines_jump_forward.py,sha256=iZWXeR3gNYoMubLGyFmLPO4V2YsN5DiGjD71Xk9iFaE,6418
54
64
  sglang/srt/constrained/xgrammar_backend.py,sha256=W7_qyyQiOUwejIPCnWgJrp6ka5fy137SiJtxt3VNruM,6220
65
+ sglang/srt/disaggregation/conn.py,sha256=amOujTy2jFwdfYhGxuTLAMNWGPbIKGoAwWDqKxZ06gc,1950
66
+ sglang/srt/disaggregation/decode.py,sha256=5pgXeIQBBJXQpVXpRm7vAauRmy5DtIi8953dZFBAPeA,18075
67
+ sglang/srt/disaggregation/mini_lb.py,sha256=upwG_4CdurUUiPuYGUO4OJQu22lDx6gnsM0xKv5QRmg,10692
68
+ sglang/srt/disaggregation/prefill.py,sha256=zw8hDy6Txq_MpC5j0fndLNcKoypT2BhxTkqqTuilMCE,9053
69
+ sglang/srt/disaggregation/utils.py,sha256=ebOZ3lSFVkbNtl6uUfS6sYYYVBjgmWdQLOsqIZBGgN0,1088
55
70
  sglang/srt/distributed/__init__.py,sha256=jFOcyt-wFAPMBUAf9zkZalNQlt-4rqmT6pCKBz1E4qo,149
56
71
  sglang/srt/distributed/communication_op.py,sha256=IBnFUdMftK_VSTMMMitGveonorFUUVNL4guqO31cMSc,1130
57
- sglang/srt/distributed/parallel_state.py,sha256=lgsscyYrqJLvblF6c5JdfYQrT-YfOKBPCHZnhazlD28,49614
72
+ sglang/srt/distributed/parallel_state.py,sha256=Lxgioo3dgqhmoyJCAiRC00GSY1elsKMyYC-tnse9vv8,49865
58
73
  sglang/srt/distributed/utils.py,sha256=U-BSaXYjWwnfG8g-tUfBhjKt5Ug097nyHtu3g3aea_Y,8473
59
74
  sglang/srt/distributed/device_communicators/cuda_wrapper.py,sha256=3jvPG-Ow5UBLiXhfx8T8snR7crSZbPpARAggsDPWq7k,7038
60
75
  sglang/srt/distributed/device_communicators/custom_all_reduce.py,sha256=kmbAV_W9ObjI7NH8X0gV__D9iHCuj6QDyhU5ZKhEQD8,22209
@@ -64,42 +79,47 @@ sglang/srt/distributed/device_communicators/pynccl.py,sha256=G-Dut_QJHOUG0j7--Zq
64
79
  sglang/srt/distributed/device_communicators/pynccl_wrapper.py,sha256=LblisImY9d6EMz-oPS9J16WHo2Q_SRL1DtlJKK63Hfg,15349
65
80
  sglang/srt/distributed/device_communicators/shm_broadcast.py,sha256=bbruDIM1GgKIdB6gi71_I0mpB179I-qyvwKuSj1Kaic,20816
66
81
  sglang/srt/distributed/device_communicators/xpu_communicator.py,sha256=ajW6132BvA6jkeipEIgN27TFycI0U06Ih2Z8WNjlA4s,1593
67
- sglang/srt/entrypoints/engine.py,sha256=xZmo4F6OuwjOV2CXQwmUAjpoj5tZTV7sJIEEXTFhZkg,19763
68
- sglang/srt/entrypoints/http_server.py,sha256=lb5sFSBKYiI7NZeFPeFeypjQkhMm5hkYqIwo9AsFDZQ,24724
82
+ sglang/srt/entrypoints/engine.py,sha256=b4JmlvK4M2cWfcbc5NOlR7HjkNzDENlzL4rbSaPjBdU,21295
83
+ sglang/srt/entrypoints/http_server.py,sha256=C7zo4C-cX84RM20KpvLzZS430fA88ei3GuAV6Clm5kw,26947
69
84
  sglang/srt/entrypoints/verl_engine.py,sha256=ICo0F860JvoCy3aKTTLpEHqKcDX4WIN5V85q6873IxA,5798
70
- sglang/srt/layers/activation.py,sha256=f097ndEIQ-lQ5JLa4HrcjqLLBeRcZbjYmj2gmVknnkA,5993
71
- sglang/srt/layers/dp_attention.py,sha256=nllN2eCd4KtiOmWdG6dQRlI3CVMi_CBzn7UihlXNLi0,6547
72
- sglang/srt/layers/layernorm.py,sha256=DI2Ih34bLh5Ex_yd8-X8L7kNRcYStpc9XpIS4zIoM_U,3860
73
- sglang/srt/layers/linear.py,sha256=8HyDjnpUu1ZxdqVRt9jiuO4GvTr2JsO42jwgtl10Qjs,51063
74
- sglang/srt/layers/logits_processor.py,sha256=BcoVrVYndmP9o0lFXj3YjZ8TOzloHXBAzLwjFZi4u-8,23207
75
- sglang/srt/layers/parameter.py,sha256=-jaiI-dIhiSPigwcQGydBiKEMSVg7XuMcg8Bzb8BHMo,15162
85
+ sglang/srt/layers/activation.py,sha256=1ykXZO0BGz7DFVE-EK26b02I5AgH2IuU4PQB6oUcF4M,6003
86
+ sglang/srt/layers/dp_attention.py,sha256=0kX1L-N_oPISkG_OdU_fcehrcXCv-bsRsFyFM_BK7OM,7279
87
+ sglang/srt/layers/elementwise.py,sha256=y2mQqjbF2FmFtNYBk5ecTyaj3ELoZyz-rWPY8rrxCtk,13765
88
+ sglang/srt/layers/layernorm.py,sha256=189bORMggKhYcEYEjl6JRcuIoUPllHo3SheoH6YiORY,4546
89
+ sglang/srt/layers/linear.py,sha256=HYIGxpRYL6x-jNOkyNtGAw5Ak9Nq8jkntddgTBER_1w,51486
90
+ sglang/srt/layers/logits_processor.py,sha256=Vp8ibljVEezTr54xzeOcjiJR7JdYO8ItkO5nLIIMVu0,24206
91
+ sglang/srt/layers/parameter.py,sha256=0OTMtmsNds42e3z3wHTRJiUfxCWFwSL6DHrqgeTgGt8,15151
76
92
  sglang/srt/layers/pooler.py,sha256=rj2lygvleBnyLCBZ8I11HGMgpfIDsT0l3PIkshJwdu4,1606
77
93
  sglang/srt/layers/radix_attention.py,sha256=UDL0y4Zasay_Rk-_XmIU4kaGbaF26ONvEHX5EQzLrqI,2260
78
- sglang/srt/layers/rotary_embedding.py,sha256=3DVmZ9UUYXPiWVxE9Ay25S0g-Oeh-_u7PXvPkRJ8nnA,44032
79
- sglang/srt/layers/sampler.py,sha256=NIHFR9oot9q-4n414zS3K3oLAHXtuT27mfv-3KgH7Dc,11638
94
+ sglang/srt/layers/rotary_embedding.py,sha256=srAyHlM18hgEQKW7MotYKt1xGkqA_j1PU8g0PO5t3lA,42770
95
+ sglang/srt/layers/sampler.py,sha256=yipSyN5UWGwGS-BC-WzWMmelys4CCDtK_8b1OpaK6sM,11622
80
96
  sglang/srt/layers/torchao_utils.py,sha256=Ws24FdRBSkTpyeyA6bQrdDm-W5wfDxKvSIPUSahyMfA,4063
81
97
  sglang/srt/layers/vocab_parallel_embedding.py,sha256=QUxd4sELx6p3dHvEKmccPZ-phdd_9EjNdwjH3SJ9zxI,22238
82
- sglang/srt/layers/attention/base_attn_backend.py,sha256=j6pDIDuxlWVmc8GzNxcAFi1USRo4gqqI7G65XAQ5u5M,3263
98
+ sglang/srt/layers/attention/base_attn_backend.py,sha256=X_GIbQuU9njtUEGdUP7E_KRhmGxj3UyPHNESlL3QaQ8,3264
83
99
  sglang/srt/layers/attention/double_sparsity_backend.py,sha256=2ZRL_gYz14idoVqQzeQ6N77nXer0f_8_TUYw40XUUz0,9161
84
- sglang/srt/layers/attention/flashinfer_backend.py,sha256=iSJ7wSlKWB2GmumG1WtakhPOjQmcI60wz5ZBeiayIHU,45980
100
+ sglang/srt/layers/attention/flashattention_backend.py,sha256=tfrMo429p0SJ4wp7ggL0_oh3J4i_2jZ0NKBLiGPZ-TA,11411
101
+ sglang/srt/layers/attention/flashinfer_backend.py,sha256=m1rOYGr9uaxeUPA8BGPGGe8yicLJKNmRm4zxkJoQU9k,45980
85
102
  sglang/srt/layers/attention/flashinfer_mla_backend.py,sha256=BgbGK5iROulOco7o3vYTaPw1Ei3EaQj6Cn9W57Z4amQ,30360
103
+ sglang/srt/layers/attention/flashmla_backend.py,sha256=iWzxEEIJ6g3RrovSReqAZMjjcv36qCIsPEHZGs_J8JI,10543
86
104
  sglang/srt/layers/attention/torch_native_backend.py,sha256=KABmBrMqKa4x08kkQYdIcZUGydvmaVJIUfo3y8jhFHI,9270
87
- sglang/srt/layers/attention/triton_backend.py,sha256=CiDI97Yp1mPOpeObpr2F8z-0OU8EPf4mQ9sQzK-8lJE,21520
88
- sglang/srt/layers/attention/utils.py,sha256=gwZoJDyJ9OIPMHWl6r3qkuyVp4Sji9juX7Pwvh9PNxI,1131
89
- sglang/srt/layers/attention/vision.py,sha256=4DjJfK4aUF994R2c_R1HQ6Q7R_CrWIyBT9NGR0yQsvY,12471
90
- sglang/srt/layers/attention/triton_ops/decode_attention.py,sha256=OcN_uL6dwAU3ImCqDHH3ZamJCQSofI8C7erTBc-ZPps,17926
105
+ sglang/srt/layers/attention/triton_backend.py,sha256=cyxOaUU1CNhaEezJH9j0dd20cwxwIVGGN3jNXFTVkIY,26714
106
+ sglang/srt/layers/attention/utils.py,sha256=J9mA-cbZT3uTlaKXo0HEAaeMei_TS2o4McTna9LVDCE,2750
107
+ sglang/srt/layers/attention/vision.py,sha256=hUUkMyhl2WbKAY8ykvolBWgIMHuihiZ7w_UkJVxUQUo,11890
108
+ sglang/srt/layers/attention/triton_ops/decode_attention.py,sha256=DPu_aCPgwPqKWZPEQmp_xA7MPbpV2ip-MEICCB470Ao,19120
91
109
  sglang/srt/layers/attention/triton_ops/double_sparsity_attention.py,sha256=BXUY8ARHBF2s9x9waiEwfZwcMgvuaJA0gxb4OeUZ_tY,31167
92
- sglang/srt/layers/attention/triton_ops/extend_attention.py,sha256=SZezMpXMr2AEEIvWBQYzx_Gnmr2fyFhQET8eucdmfdU,13244
110
+ sglang/srt/layers/attention/triton_ops/extend_attention.py,sha256=hbBvnhc2zqu-E3HNROVXyNOZbtDkVRuFus-yTjmE0Sg,13668
93
111
  sglang/srt/layers/attention/triton_ops/prefill_attention.py,sha256=Y66gZ37u0GKMPtI8n5MbO6uOxRuGEmKIG0IPbJTOqAM,6213
94
112
  sglang/srt/layers/attention/triton_ops/rocm_mla_decode_rope.py,sha256=664WnAJ91EiCUZOcnVDfbTQf4uGJ4ZDZB1CbxpEUFZc,13866
95
- sglang/srt/layers/moe/fused_moe_native.py,sha256=domK1jDc5-zJcdwDso-YgBp_O136VHifnvFVlfFvKsY,4325
96
- sglang/srt/layers/moe/topk.py,sha256=MOwBNBwxxW__mgmpX08RIqrh77aXLEgr0F5b5iF7hRI,7512
113
+ sglang/srt/layers/moe/fused_moe_native.py,sha256=KF0J5W5biWYWbERUHAduQFnUV5om9VaYv0I9avsVfgU,4330
114
+ sglang/srt/layers/moe/router.py,sha256=gvyK7hXlujfCZCmAIFc3oxfgjuAjzlpPe3mp1Blc6Y0,10419
115
+ sglang/srt/layers/moe/topk.py,sha256=0hST6zPHVC4_nPMzQGtt-POlwtk3aqj37EUKlFXTvQ0,7836
97
116
  sglang/srt/layers/moe/ep_moe/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
98
- sglang/srt/layers/moe/ep_moe/kernels.py,sha256=hG7E29bbj8HeXEkAd6N0HzJZPn7cwVx1VJptzpRWOv0,15376
99
- sglang/srt/layers/moe/ep_moe/layer.py,sha256=y8YaVIvzZ5bomc_9wvc1SwcgOddmVPlD5FYQfxI1kPY,27890
117
+ sglang/srt/layers/moe/ep_moe/kernels.py,sha256=E-VyToHN13f443U3UboDgLwv-t8gAcXOtPHkrI2xUpI,18793
118
+ sglang/srt/layers/moe/ep_moe/layer.py,sha256=ZiS9viPdcpB7SjylNVB3DSuF3Az1jGjDPdCqd657Qf0,36990
119
+ sglang/srt/layers/moe/ep_moe/token_dispatcher.py,sha256=SPHfAXIkZmFqXMU0a2eob9j_cni269v5jevGm2x1vWM,14845
100
120
  sglang/srt/layers/moe/fused_moe_triton/__init__.py,sha256=h9yMFAL_bagUf-qBED8gSWdCOb7d8IdA-pE-L_nIg8E,842
101
- sglang/srt/layers/moe/fused_moe_triton/fused_moe.py,sha256=R1XYyB-oS0QMxIoGrMKPBlk17Jq-z1dTzafZ7bcTSwU,41856
102
- sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=tHSxJ2JWwaSwPuyleILJiMH4cNrjf9P5Y8dcAJ8gm5o,24129
121
+ sglang/srt/layers/moe/fused_moe_triton/fused_moe.py,sha256=lBF40JXOgUzFMza020gRYol4WdLME9OB-YA9tSfsJ3k,41516
122
+ sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=CYT3ujh5ifonhqQc1uYSa6maJADUOBJoRHMIWmf-DU4,24100
103
123
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=1,N=14336,device_name=NVIDIA_A100-SXM4-80GB,dtype=int8_w8a16.json",sha256=iNGsE2ZeVnQEnN4A8UJ9Jv0d3hbRF2MJ9oBgjup5Szk,2737
104
124
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=1,N=14336,device_name=NVIDIA_A100-SXM4-80GB.json",sha256=JJN0hryyLr5Zv3dSS7C8cPFhAwTT6XxUVnBGMZvV6JA,2752
105
125
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=1,N=1792,device_name=NVIDIA_A100-SXM4-80GB,dtype=int8_w8a16.json",sha256=ouRyZ5PEMPP2njPftCNhs-1g1y6wueWLmhI7G1SjV1k,4131
@@ -144,6 +164,9 @@ sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=tHSxJ2JWwaSwPuyleILJiMH4c
144
164
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=256,device_name=NVIDIA_H20,dtype=int8_w8a8,block_shape=[128, 128].json",sha256=yJzHX8nNWnKwj9M8281dwwPhOHQ039mBpvNgG7poy3Q,3264
145
165
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=256,device_name=NVIDIA_H200,dtype=fp8_w8a8,block_shape=[128, 128].json",sha256=R4B2n2vGt4pPo6jS4Bmnx8AYtcfF9qQJE5bD7OhmXHs,3265
146
166
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=64,device_name=NVIDIA_A800-SXM4-80GB.json",sha256=Y4WpmVGlGERHoeoQNGkQ-GC-MsEtMblqnAVuDbARJdw,3240
167
+ "sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=64,device_name=NVIDIA_L20,dtype=int8_w8a8.json",sha256=RUkd9fW9WbajF_fFIzppsE1qyWGR5aRC4Cln-BPdu28,3254
168
+ "sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=64,device_name=NVIDIA_L40S,dtype=int8_w8a8.json",sha256=Sc9xK1wtRUqIzXppbutcq-Y2e9M0DZl2OGVzzB0aQuI,3265
169
+ "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=1024,device_name=NVIDIA_H100_80GB_HBM3,dtype=fp8_w8a8,block_shape=[128, 128].json",sha256=TBscr1uWwpY0FrKQ5Y3EO_Qg6I97u4f_zjnWRvoeLvE,3260
147
170
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=1280,device_name=NVIDIA_A100-SXM4-80GB.json",sha256=6QPLhZurIqcMVdy3w0Dd7gLViKxsyJRBz-qd8agpi6Q,3248
148
171
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=1280,device_name=NVIDIA_A800-SXM4-80GB.json",sha256=uv-RwTNZT2n264dLo4eWxUpB3g7QqUyf2MFEGiRvoqQ,3251
149
172
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=1280,device_name=NVIDIA_H100_80GB_HBM3,dtype=fp8_w8a8.json",sha256=XbCRIOmiNqVKh89p-0UxvvspINRDA1iV83f9l5yORwA,3254
@@ -157,6 +180,7 @@ sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=tHSxJ2JWwaSwPuyleILJiMH4c
157
180
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=320,device_name=NVIDIA_H100_80GB_HBM3.json",sha256=prj-QxdfS9Ns5WRPvahY_Tr7CyqlaVgNHPT89SS5zzg,3239
158
181
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=320,device_name=NVIDIA_H200,dtype=fp8_w8a8.json",sha256=AUgoHK1PmAFehSNmsbxunlBdzM50Q5nFvdnG9FSOjOw,3265
159
182
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=320,device_name=NVIDIA_H200.json",sha256=pC5fdtEFc5aVNzpj_REHhz1QPrGvgI9iQCvlodDP7J8,3244
183
+ "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=512,device_name=NVIDIA_H200,dtype=fp8_w8a8,block_shape=[128, 128].json",sha256=yZ_6KgwJvKPGblpbHPQvEMcUc9F7G9A0mSXHfgfztgo,3263
160
184
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=640,device_name=NVIDIA_A100-SXM4-80GB.json",sha256=oxOKFDrgmw1YmgxTtRa1uoe3p09ylTLrkj_jOTqNh1Q,3249
161
185
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=640,device_name=NVIDIA_A800-SXM4-80GB.json",sha256=bfr70r6PmM95w7raabQOaOOSPiwU2OQCOZh-kKXIehY,3248
162
186
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=640,device_name=NVIDIA_GeForce_RTX_4090,dtype=fp8_w8a8.json",sha256=4g9lABUJnB-iVwXfYPYcI05XFPG4jY8o0yJUK7kSPZM,3253
@@ -215,18 +239,27 @@ sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=tHSxJ2JWwaSwPuyleILJiMH4c
215
239
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=8,N=8192,device_name=AMD_Radeon_Graphics,dtype=fp8_w8a8.json",sha256=-RzUWSIAAsg6iA-8SPMa68hPpBVoUyMJs3dLP7edRu0,4323
216
240
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=8,N=8192,device_name=NVIDIA_H100_80GB_HBM3,dtype=fp8_w8a8.json",sha256=sY2nWMPh9lsIkhPCjkHO245wpnfFbrHmzdcZDVFPVww,3265
217
241
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=8,N=8192,device_name=NVIDIA_H200,dtype=fp8_w8a8.json",sha256=Uz5X80VcNBOaxshwVNUEittHk2zqB4HQCfTJ4TPG5aM,3274
218
- sglang/srt/layers/quantization/__init__.py,sha256=eBbrWZLYhjtzva61n0VTvs4Q60hb2S8qlHRuiPSFiME,11086
219
- sglang/srt/layers/quantization/base_config.py,sha256=daK9p0aijMszLUm1W4Pc33FK87MdqYK1NoWFKif-j80,4599
220
- sglang/srt/layers/quantization/blockwise_int8.py,sha256=AhMwTDFkE6NmxO_ER6rq1K3k5Bo5ZZAFk1OmEju7rEo,14908
221
- sglang/srt/layers/quantization/fp8.py,sha256=pX94MLbzeDuwSgagKux8MInW4oeZ0YEm_j4A6mDq62o,39758
222
- sglang/srt/layers/quantization/fp8_kernel.py,sha256=BNQC6MlDGw5HnaSBfmn7vajoLEE57J8boAy_2KmJ7DU,23816
223
- sglang/srt/layers/quantization/fp8_utils.py,sha256=IOhwQv7eDT0-L2H9gGOPx5gu7lGJxUHDubltAnNU5ps,11100
224
- sglang/srt/layers/quantization/gptq.py,sha256=u-WadiLnVbeskCU8Ar-1IYofraOf1caYkGrUFFYC27k,14651
242
+ sglang/srt/layers/quantization/__init__.py,sha256=j2VNj7XgeaM_J6aWGjee6ECOojxEFkp78wDugM9mSMs,12834
243
+ sglang/srt/layers/quantization/base_config.py,sha256=jWk_egQrVNMYmQgbTI9vkcgzScLFjB5_sywFlAfE5J0,4776
244
+ sglang/srt/layers/quantization/blockwise_int8.py,sha256=BS4nmo18QaC3vyCe9Wwe0Te-7FmFUe-udl7xbhRaU0s,14887
245
+ sglang/srt/layers/quantization/fp8.py,sha256=J9P6SwZ1PIb3dEJoP7X2Uw5VOtU35vU5jugkoYyaZC4,41081
246
+ sglang/srt/layers/quantization/fp8_kernel.py,sha256=_DwqmdyJ4hObRgdOGByn0IAbwUpzrjOxtgPzQ_TjsJo,24349
247
+ sglang/srt/layers/quantization/fp8_utils.py,sha256=OL_tV-NvNGcXV2i3mOXytjzi9d4t59vOxvZN_hBc1KA,21572
248
+ sglang/srt/layers/quantization/gptq.py,sha256=IBQOAFKK98_LsgU-iuYjnDtmOnm20hRniI0VCQGDtxg,15228
225
249
  sglang/srt/layers/quantization/int8_kernel.py,sha256=GfRn_imIw8kNgqdtb2lr7BettjgDgimbl1Rubnamjh8,11352
226
250
  sglang/srt/layers/quantization/int8_utils.py,sha256=YK9CS-lb_n91kNCTKK5o5apYF31V2giDg5G5VKrpcUA,2356
227
- sglang/srt/layers/quantization/modelopt_quant.py,sha256=F16S-ky8EiqBrTU-u3gCmyc__TVPu03bR1Xo8ojCzRk,6989
228
- sglang/srt/layers/quantization/w8a8_fp8.py,sha256=AOczkwzh6gRx6nvYFqbrdNyAxmlPsXhws1DEX_nP0Ss,3672
251
+ sglang/srt/layers/quantization/kv_cache.py,sha256=rJi6amyLZsquUMo_V5iLlPMqdsGTLgxh4popN1xUHCQ,4236
252
+ sglang/srt/layers/quantization/modelopt_quant.py,sha256=mne4uKF0R-K0OvWN7X5ZxD4LdXKBc6GvmpZzIW6gkmM,6969
253
+ sglang/srt/layers/quantization/utils.py,sha256=BmbovB_SxPcV7DRzGod8plawLbWRLIiN8Q2l_oWtAfw,5627
254
+ sglang/srt/layers/quantization/w8a8_fp8.py,sha256=jw0oGs0-0q4pu7aG3YtSZnBnO6kFlCDiNOTfij7xo2c,6233
229
255
  sglang/srt/layers/quantization/w8a8_int8.py,sha256=V5vxn0wmUL1szj38lsJOKeNNEvFHisU7hZZLO4FfoNc,8733
256
+ sglang/srt/layers/quantization/compressed_tensors/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
257
+ sglang/srt/layers/quantization/compressed_tensors/compressed_tensors.py,sha256=YEtWykakVRjS4rko8T7ui0uf_Q15n9SPcASY9EkWYGk,25344
258
+ sglang/srt/layers/quantization/compressed_tensors/compressed_tensors_moe.py,sha256=jHuFgm6s0xxijtceDySj-myyWVn2JJMx4ZXn-4P16F4,24985
259
+ sglang/srt/layers/quantization/compressed_tensors/utils.py,sha256=mnUmKWFQUnY8bVoFHUuNVwqsfS-cefeR-ofyaihCXcY,7621
260
+ sglang/srt/layers/quantization/compressed_tensors/schemes/__init__.py,sha256=qcdRgoUNJWXqSimns-D987TW0OTk3uFuWNBX8Z6H8Fk,246
261
+ sglang/srt/layers/quantization/compressed_tensors/schemes/compressed_tensors_scheme.py,sha256=tdKJC8c3SX8T3z8JL-1YCsg4ftcv55Wxt0vZrYftpX8,1635
262
+ sglang/srt/layers/quantization/compressed_tensors/schemes/compressed_tensors_w8a8_fp8.py,sha256=BSgSFFeuBHSnZxvVpNDojeGCx0ClDDH04aTjdBUyE6U,5940
230
263
  "sglang/srt/layers/quantization/configs/N=1536,K=1536,device_name=NVIDIA_A100-SXM4-80GB,dtype=int8_w8a8,block_shape=[128, 128].json",sha256=RdHQxWXwXqvio31192vsLaKjEr4f_DjpMPKlarY1IAk,3251
231
264
  "sglang/srt/layers/quantization/configs/N=1536,K=1536,device_name=NVIDIA_A800-SXM4-80GB,dtype=int8_w8a8,block_shape=[128, 128].json",sha256=0vLaJgo5B9ti-XMFKJuvSoMGjsZQ-RhHSx4cC8Xji-U,3254
232
265
  "sglang/srt/layers/quantization/configs/N=1536,K=1536,device_name=NVIDIA_H100_80GB_HBM3,dtype=fp8_w8a8,block_shape=[128, 128].json",sha256=tkLjwLC_aVXhzuvo-2QHkojXZauPJsf3jNHFn1S7uRA,3244
@@ -379,12 +412,12 @@ sglang/srt/layers/quantization/w8a8_int8.py,sha256=V5vxn0wmUL1szj38lsJOKeNNEvFHi
379
412
  "sglang/srt/layers/quantization/configs/N=7168,K=256,device_name=NVIDIA_B200,dtype=fp8_w8a8,block_shape=[128, 128].json",sha256=PD4AJYCkHfy2ivv9baMouFXzBTy0eKMumbAfxfm91HI,3256
380
413
  "sglang/srt/layers/quantization/configs/N=7168,K=256,device_name=NVIDIA_H20,dtype=int8_w8a8,block_shape=[128, 128].json",sha256=FImA-TJ_tQDjqwoNWxS--sRDoKDXf9gamlME3tkxH58,3252
381
414
  "sglang/srt/layers/quantization/configs/N=7168,K=256,device_name=NVIDIA_H200,dtype=fp8_w8a8,block_shape=[128, 128].json",sha256=FFBjSWlpKXMxfAUUYUqXbOK_Hd7qBeBsfbcaa9uB4qY,3249
382
- sglang/srt/lora/layers.py,sha256=r34oprzwyE3SWPvaNkBvXWPtfa-0IY987_bjj36ySfw,9996
383
- sglang/srt/lora/lora.py,sha256=07-IaAfbb3zPJ-DukHL3uyQ8fjJx_hrVFHS2pqe8LZg,8238
415
+ sglang/srt/lora/layers.py,sha256=2ZlkwSs_OOLkgbeWADUb8f-LOEear0yFZTup5mNjMBI,12543
416
+ sglang/srt/lora/lora.py,sha256=9BpasJObx-XL_qSoQhheGaYBJ3PlKqJAARvruIOJE0w,7514
384
417
  sglang/srt/lora/lora_config.py,sha256=qDgMTx_69jyJUl29O5FxLzYa0BMhqYVXWXfyyVOvGm0,1684
385
- sglang/srt/lora/lora_manager.py,sha256=Wlq5dxLM7Uj4uTGpFXH1q-IOI8j4mFXYHPKSltx2QMI,7794
386
- sglang/srt/lora/mem_pool.py,sha256=eV_GXETxNODPVIAnTEeUUUVn0IVgguBR_mYFzIK-VHA,6835
387
- sglang/srt/lora/utils.py,sha256=6i7Q1Y-1LLbRkeCMv_lKIzkTN0veUTLbc8wlHn7R-bA,4571
418
+ sglang/srt/lora/lora_manager.py,sha256=pchu-IQ-SYm5G99ZSW2PORUvD4JCuljVUcQZpYcNkaE,8634
419
+ sglang/srt/lora/mem_pool.py,sha256=EEv2bInK0muq60bEdm4M8e5nCvF96cj-T4TZbIjpMZ8,9398
420
+ sglang/srt/lora/utils.py,sha256=_NA2l9WoHzFSJUKituYGbzs0pfFZMH6vukAUUUgbqlU,5081
388
421
  sglang/srt/lora/backend/__init__.py,sha256=FziFT8HguMFj-h0tUCc4_UEbtOWMlYi4gNlYJcArWh4,671
389
422
  sglang/srt/lora/backend/base_backend.py,sha256=dldwA7vTWrB1ln1MwLYKNtMkBoAgD7OLSlWe9tL2lzk,4602
390
423
  sglang/srt/lora/backend/flashinfer_backend.py,sha256=fXfkl7Cpw8ap2bCrgWdn_gEUzMXX1pNjNuiPw3kA76U,3984
@@ -394,112 +427,122 @@ sglang/srt/lora/triton_ops/gate_up_lora_b.py,sha256=qve4oNZHYUFk9ckmT2BVuDNMEvrN
394
427
  sglang/srt/lora/triton_ops/qkv_lora_b.py,sha256=BmIcTZMnlSnie9rnMl4KvLpc4Njsk7_IppbUqitf9Xw,5738
395
428
  sglang/srt/lora/triton_ops/sgemm_lora_a.py,sha256=kv-AvJ_Bi3yWjGvFnSwXvP66iJvY9n9pEnJzJ9-DWzo,3982
396
429
  sglang/srt/lora/triton_ops/sgemm_lora_b.py,sha256=Ai5vPriT4OgACwK7xrpGgf5L1oaN9x0jwNKMChu3uI0,4299
397
- sglang/srt/managers/cache_controller.py,sha256=ZfMXH3m86aW5F5TkxDpSNwk_C2e1Irdyn2R2HWTha-Q,18037
430
+ sglang/srt/managers/cache_controller.py,sha256=BRRyu110lJalLVLZbTCMOWbD4s-LG3f9Xz6sxU2dCrc,18144
398
431
  sglang/srt/managers/configure_logging.py,sha256=fOJaXAQ1n9m-8KPJndpsKvS885i69SMafoEADLIVfIM,1633
399
- sglang/srt/managers/data_parallel_controller.py,sha256=K6jwHn_UhsC7o_lZT5FQm9oKBbsYhlHKiDSutFlm3jA,9539
432
+ sglang/srt/managers/data_parallel_controller.py,sha256=Xkj2n9uDyq7a-AVDZlfzeuNkC4ibsSftb1_bed9hgQ4,10318
400
433
  sglang/srt/managers/detokenizer_manager.py,sha256=HTfpJWMF1EImhKOnLJ96xPmYXm71xzaisLMfxg3zpgs,10111
401
- sglang/srt/managers/image_processor.py,sha256=o2ChZW3mWSw1I_4HFE8SV1ZsYtWp9G9w7Ij73BvugN4,1711
402
- sglang/srt/managers/io_struct.py,sha256=XvXM6WRarwEtaNQWlzZpYKy2Zi4wCG3xnz_ojskYJjM,22791
403
- sglang/srt/managers/multi_modality_padding.py,sha256=dkFKknxML8R3KrAZhKXO-A69u3pUUI1pRfOVMJfqHYw,5165
404
- sglang/srt/managers/schedule_batch.py,sha256=FZbC8bRej8ZhzW6pLelVQ7M8dwgnbgN_i-QZ3JrcWgI,55964
434
+ sglang/srt/managers/expert_distribution.py,sha256=r3o5RGI0gnV7xb60AApqKYa0oiSB37oB7hQBX7P3xZM,3225
435
+ sglang/srt/managers/io_struct.py,sha256=RQOIJwA30hkXlVPiLv--ELdFRUjEDNzrZJ0vFsNHzJY,23722
436
+ sglang/srt/managers/mm_utils.py,sha256=KzodrStj3ouIEHLRja3TUml0YUQ59qmEPejks5ikPQk,13828
437
+ sglang/srt/managers/multimodal_processor.py,sha256=nTlktTXYo2NKr5Ab2AaSjydNQDdwpJHs4XdYr5zdtkU,2154
438
+ sglang/srt/managers/schedule_batch.py,sha256=bjqWg0tdDkqT_AqGijsNnOJ4g7cKGbO9lMYItJ7Lv8E,60033
405
439
  sglang/srt/managers/schedule_policy.py,sha256=E1qVq2G3jptKdX9nlqfayeRBUll9xB6bK8nBf3EW32E,19469
406
- sglang/srt/managers/scheduler.py,sha256=td0JxX0nKrrfgXu2wBPtDP4YHrDOsJr41XDox0LUB10,68030
407
- sglang/srt/managers/scheduler_output_processor_mixin.py,sha256=kPOocSA78LzTbbhmPoTkguIoiFOiN0gzHil_kR2aGqY,26253
408
- sglang/srt/managers/session_controller.py,sha256=YOv8cFcuVmdCE4OfQJ6aA5AosHRwtZF9WdUUfUXEp0I,5753
409
- sglang/srt/managers/tokenizer_manager.py,sha256=_xIn-3BDgagqnmRKFhIQTkCg2jXk6jp607B7CeprNVw,44686
410
- sglang/srt/managers/tp_worker.py,sha256=o9MY1a8x81nI3W0m64YvOXcAA4sCBvPPmnNU67vBXGs,8710
411
- sglang/srt/managers/tp_worker_overlap_thread.py,sha256=BhbVjI4DDXBIbRs1xzxME5uHy6WAxmZSJBepRu0Hv-0,9108
412
- sglang/srt/managers/utils.py,sha256=5i75uLlQOF_5CaT02CrWtwozMTtwTg2_nLP8Dtr-JZQ,1536
413
- sglang/srt/managers/image_processors/base_image_processor.py,sha256=6UdYo3lFgrMsn-5a8NDKhkjeJ9bxPDFRVzhXv_wV5fg,7567
414
- sglang/srt/managers/image_processors/janus_pro.py,sha256=2Qmt2gKPT4-pUnTBhEEokiBEQ0kEjdZjl0Eq1U6y42U,2615
415
- sglang/srt/managers/image_processors/llava.py,sha256=UdNfmAZuw__MgySC4uu5NsWgBXPyP4V4lIzDEFV9AHs,5839
416
- sglang/srt/managers/image_processors/minicpmv.py,sha256=JS11rO_gyRVLijD2AxNM9ZJZOFxLbver_v3bqqUEhBg,2881
417
- sglang/srt/managers/image_processors/mlama.py,sha256=ECN2rmS2Vl9eg3u-XmwtphQm7zTaB4clK2HU5ntuODE,2186
418
- sglang/srt/managers/image_processors/qwen_vl.py,sha256=wJbbCW1DKPQp_RufQ4fz7Nd_hRldfwe97e_2YMUGmJI,6417
440
+ sglang/srt/managers/scheduler.py,sha256=jQn0IZTH4Ri6d21ngj-5na7Pz4sSR3OfEfY16OxcKzU,79713
441
+ sglang/srt/managers/scheduler_output_processor_mixin.py,sha256=u2sj6MViFTov0lVZSysZ-wph2pEqRCtCjwA1UdttZ7I,26338
442
+ sglang/srt/managers/session_controller.py,sha256=o-ifit0n4_xHLNmyD0Ams8FxGRgxFybX-Vz1hwgr3UQ,5755
443
+ sglang/srt/managers/tokenizer_manager.py,sha256=6Va7_jPK0AeEnpzODT7JePSDjIzu6KDo1z70USwcS6E,46244
444
+ sglang/srt/managers/tp_worker.py,sha256=pTLP-RkLfYempJ1bHzjpY0yQRyhGBsBTtIAlOsVaH2k,8724
445
+ sglang/srt/managers/tp_worker_overlap_thread.py,sha256=3_ZJ8Rq7v2ZDaRNTRu5Dy8AbqiAlJQp3IAKnn_WAwd8,9127
446
+ sglang/srt/managers/utils.py,sha256=59IuYvuEfi8ji_acKqH3Y1-2PaA_dSlZMbHSCqjQfCA,1629
447
+ sglang/srt/managers/multimodal_processors/base_processor.py,sha256=PpNwQBhEuXbztzS3cBFu7UL5sfKYay-WoqAweSVRRao,9984
448
+ sglang/srt/managers/multimodal_processors/deepseek_vl_v2.py,sha256=-qU99yCs3TYM8g9vOo5_SIWX91theu0ZREXZT5-A9yY,4594
449
+ sglang/srt/managers/multimodal_processors/gemma3.py,sha256=Ic8wTOS5TS9qc-SsGkKWQ_7NgnimCnnBeXw4UB8DTgM,2745
450
+ sglang/srt/managers/multimodal_processors/janus_pro.py,sha256=GaVH3_HxGcQHRvFuP5zpkQ9m678cyVrbjb1iwwdV_fE,2780
451
+ sglang/srt/managers/multimodal_processors/llava.py,sha256=2UfMVl-CoQ00Wt8XCZLkE3G7mFdU4Ol1aid6PPG-ZXw,5711
452
+ sglang/srt/managers/multimodal_processors/minicpm.py,sha256=eD-FgRMTSF9jvolMQYfIgzt31qs0SyLEcVrpHaf46fU,5979
453
+ sglang/srt/managers/multimodal_processors/mlama.py,sha256=K-_guz4aaxR8R2zm9MAnLy7K66fkd_0J6R9P_4wdIb4,2118
454
+ sglang/srt/managers/multimodal_processors/qwen_vl.py,sha256=OBSKa6LGhTl6wJJ24Bk0IibcKM97Q-6XeBJ1jxLGMbY,6654
419
455
  sglang/srt/mem_cache/base_prefix_cache.py,sha256=NY62Zo0A0tLJ7ObRLOQqQcXCxoJUDZsK8f5U4dNQjKc,973
420
456
  sglang/srt/mem_cache/chunk_cache.py,sha256=it5SfL1FwMbrdeOH-I-Eu_i-I9hFB1xL-z_brIUoCkk,1835
421
457
  sglang/srt/mem_cache/flush_cache.py,sha256=GYcxmNXh4hsMpFfNOuCTpKilW7guZwTtAg_usVeM3J0,979
422
- sglang/srt/mem_cache/hiradix_cache.py,sha256=qBC-VGtkS_q8g0KoeYWIAQpf1fLJvMf88GHcRl35nMk,15300
423
- sglang/srt/mem_cache/memory_pool.py,sha256=IayAcOXRWA6IFOVxVybXEC-UgSmXNsOSty7KoIRKLBQ,22681
424
- sglang/srt/mem_cache/paged_allocator.py,sha256=DRq7qZXLVeHKkCWRlEkoAmLP5IC_1ReHa_xL7-8Orzs,8594
425
- sglang/srt/mem_cache/radix_cache.py,sha256=bLldgkFZvmFOVgc2xLI87an3UBYxIsOlRBsxenFEL7s,14912
458
+ sglang/srt/mem_cache/hiradix_cache.py,sha256=BjPZXHYmCW6Sdohb67Qqzf2T0YXkO0hEPyynvn65xHs,15956
459
+ sglang/srt/mem_cache/memory_pool.py,sha256=xfEFCYs3DsW7t5Nuv9pwFKtap3SJgxkvXToKQFUT-rI,28255
460
+ sglang/srt/mem_cache/paged_allocator.py,sha256=Fl6d8rgkwGIgq3n7AKM7Jmm_aIgu86jJ24636nIkX6s,8594
461
+ sglang/srt/mem_cache/radix_cache.py,sha256=Lm-pco6CJ4orb9IfDpbHm5MnyK8Ya0OF1x9p88dv548,14906
426
462
  sglang/srt/metrics/collector.py,sha256=0X40ZZ18182sx2t0eqeqoK7gspH36L343zNvSkgBvd0,9293
427
463
  sglang/srt/metrics/func_timer.py,sha256=VFyNRrbnKVCwnQsrlLin1lITJfjQpf9m8sGPqL5LIsQ,3438
428
- sglang/srt/model_executor/cuda_graph_runner.py,sha256=-zhSlRHgXt3aTP1XToKmFu2645-ezczayBrilmxTo0g,20399
429
- sglang/srt/model_executor/forward_batch_info.py,sha256=rT7YaTlycyF4Rjm6_W-g0Hcqk-XwDLfDljricK4Q7_k,16992
430
- sglang/srt/model_executor/model_runner.py,sha256=zE3nWMBLI0eW2mDZHBvxj1bHW3VsAxtd-RP6uddyP5A,41581
464
+ sglang/srt/model_executor/cuda_graph_runner.py,sha256=H93qEHIbC7SukxnlgaswlRUhY6b-gh5lm1Je2C39rD8,22880
465
+ sglang/srt/model_executor/forward_batch_info.py,sha256=RKRQMoMcAexG-YweQKl-uSXc4qe-yk4u6iTtWK_BOpw,19591
466
+ sglang/srt/model_executor/model_runner.py,sha256=_86C2eQQulrBS7bV7QTibeTy29iKfaQIeKWARtZLRCo,44661
431
467
  sglang/srt/model_loader/__init__.py,sha256=zGZkOBz1zx-pkaIy47BasL3fjDlAcxAXUTjInOhXHAE,919
432
- sglang/srt/model_loader/loader.py,sha256=bHsb3T9CGac32NgDeRt03cJXBy-t1WsyYOWcYkeMVLA,46813
468
+ sglang/srt/model_loader/loader.py,sha256=FMZPejZIRjaaG9pGnjLSeJ_Jg5WfBJVT1gzWKwH-2b8,53586
433
469
  sglang/srt/model_loader/utils.py,sha256=0NaMR67fESFopaklmsleiL27XH1QUrjZW246MUu1EJ0,1369
434
- sglang/srt/model_loader/weight_utils.py,sha256=26fCuVipwEiDjziwsuavwa_BC1zPB2-socfly6IjmG0,30373
470
+ sglang/srt/model_loader/weight_utils.py,sha256=IK7Z6biH-04QZazPfCZMOMx2TXBIVdzOtb1h7LLrn2w,31996
435
471
  sglang/srt/models/baichuan.py,sha256=iXgta-W38OWesxmXWZJ73fUvPdu51EwTQzUD5mmfJ8s,15721
436
472
  sglang/srt/models/chatglm.py,sha256=avLC7mjjGskBxCxy-9s0sMlAJjfFoG_y8VieR1QfDsM,13918
437
473
  sglang/srt/models/commandr.py,sha256=Ug-B0QcdWZufrTybC6K5yP3MLKNsYb-vzfrqUsXYUcI,15276
438
474
  sglang/srt/models/dbrx.py,sha256=0Vf4yhqe8YeQuKR3P-agvYJScmHwH3-tFbyU8kv5QJM,15559
439
475
  sglang/srt/models/deepseek.py,sha256=Le2MXij8m4hT7QYgD0bFMFmYhbgX7SMjoXZFB8BxgyA,16871
440
- sglang/srt/models/deepseek_janus_pro.py,sha256=sduOYeAQrb2DFfoeJVzxMNOAgS7_CmTqaKWc8J3Ypeg,72528
441
- sglang/srt/models/deepseek_nextn.py,sha256=8uR4k1qCBKL1JAcE8cCJz3HowrIgCgdMtbqQ4Xq8Tr4,13161
442
- sglang/srt/models/deepseek_v2.py,sha256=OgazDSyX4xp5AfDBulLXOy2yGk5lwAftqNY5STLCQNE,48984
476
+ sglang/srt/models/deepseek_janus_pro.py,sha256=-XwO_3-NZhNYeCPVpUJyKtGP4TxIvafSbutl-lh_aYs,70527
477
+ sglang/srt/models/deepseek_nextn.py,sha256=kca-2Fm2_SmqbOEFfd80pobooi1BXd1oe_4EsUM6SeI,13561
478
+ sglang/srt/models/deepseek_v2.py,sha256=IqqNBO3QoJx397pBfAzD5cfqcGxBT8iWXeO4Cx0b3QQ,54139
479
+ sglang/srt/models/deepseek_vl2.py,sha256=T7zdLBnnBSVPCX4cec1B-MbnXru1-70KqEdDLqbYb_g,12956
443
480
  sglang/srt/models/exaone.py,sha256=5iibqQTjpgosuGRt2rj2lWR0ShK2XGhbdFSnOWpaQss,13386
444
481
  sglang/srt/models/gemma.py,sha256=3XxMDOKz4xMP6VzWoW8f0hmMf8LP8fhzMw5prsYC4e8,12602
445
482
  sglang/srt/models/gemma2.py,sha256=MDe_HNkSpEJpw426tbx3fp271GBlSVEuhIdGeOB_jYA,16356
446
483
  sglang/srt/models/gemma2_reward.py,sha256=V8U3_ADUHWPdOwvEe1jhGW-oJmBgL8t1TY3-67Ksv2A,2618
484
+ sglang/srt/models/gemma3_causal.py,sha256=mCUJqF5LvCpaQgeRkoR6wV33vP_7wVfyxBtKTmsmfiA,24454
485
+ sglang/srt/models/gemma3_mm.py,sha256=KxZ-uFqkRJ58RexZAGd7aw9aw2Kb5EI0DBJ4mHphqoQ,17630
447
486
  sglang/srt/models/gpt2.py,sha256=dAnfmsAL7JVHakryqrERR1jgL8mI1Op6nPHYfDCF7Ao,9802
448
487
  sglang/srt/models/gpt_bigcode.py,sha256=EAN6xAXpa8m3DcBuH1D4rTPji2oG9NSozGXSNHtE2lw,10268
449
488
  sglang/srt/models/granite.py,sha256=nu_Zl_PYn188gk1uYVZ76y4wwHZV7G0w7uanhqpSFUs,20813
450
- sglang/srt/models/grok.py,sha256=LYi-1VpGiB4SvFBc3Scp7vQTjiCODa6J_bFMjSOdsCQ,18768
489
+ sglang/srt/models/grok.py,sha256=pQOXtpHOYVntwt5QQRLffYsnMHmMfPMmGyKMfR0k0Ic,27994
451
490
  sglang/srt/models/internlm2.py,sha256=4eh9WVgK4yg13IsnH5qB2xUCWnixj_aLLz7qa_4m2_Q,13017
452
491
  sglang/srt/models/internlm2_reward.py,sha256=ndfGmyqYZbVZ7C7rJ-v9oK3wa-EpoBGybS8MlyKZi2E,2522
453
- sglang/srt/models/llama.py,sha256=FEtMrzjrbQLOOQvrqGFwslizjHPaU--9rEiXlr7MIiI,22958
492
+ sglang/srt/models/llama.py,sha256=0jD_xcCuLfDGfHrzwn3oRZvRxilaWb4ARasu5U1x9zQ,24529
454
493
  sglang/srt/models/llama_classification.py,sha256=4QWTFaUZIFKYZvEzs8bx8VkOZNIwdYCLrnwrdAw4QK0,3108
455
- sglang/srt/models/llama_eagle.py,sha256=4ynCbF2Lp-t0_T036N_Pa_8a1SIPkNeWuEwF7r5x1ls,4819
494
+ sglang/srt/models/llama_eagle.py,sha256=OB2lKsjn7BcfCZljklnhk83me8j0PuQmYLou7baNcq4,4866
495
+ sglang/srt/models/llama_eagle3.py,sha256=v3bftBVDIGjnzngQYnu19cy0J_3w7yruHqLP5nsAQDM,6642
456
496
  sglang/srt/models/llama_embedding.py,sha256=zq-_lNu35VBFc7eemiam0zdkGIE8fzrgk5OWYfirZnA,3254
457
497
  sglang/srt/models/llama_reward.py,sha256=LF2nqMV5XOrljGjAwJg43mBv3z6Q040I2EYlgZeCp8k,4681
458
- sglang/srt/models/llava.py,sha256=J06XzASrhg2Pw1Z_jMoC1lXI4hFfoZacqS2mhiLI41k,26778
459
- sglang/srt/models/llavavid.py,sha256=iwqwTruJTG9D1zV95RHq-RpTp1g12kKPrNIfv2a2XyI,12485
498
+ sglang/srt/models/llava.py,sha256=G3aaWr0Pldx-yqmbv33KMMwpJ7TAW0yIY-eE9bFfWto,26785
499
+ sglang/srt/models/llavavid.py,sha256=w_eaZRSObvqftKE_s0cAE5NvUyFtHfU1EiD0XFBfSow,12492
460
500
  sglang/srt/models/minicpm.py,sha256=-ot45U_Bv4x85JdbIAQXoxa1sF-ZDkBk8flU-Ruli5Y,14652
461
501
  sglang/srt/models/minicpm3.py,sha256=sRHPFUH636GIY94B-hpAN2MSzYT1pzLPVypTNjUtttY,26270
462
- sglang/srt/models/minicpmv.py,sha256=AYxGTFbvfhIgmGmM7ab6620jtCS2qabh5tJ6NrkQ_0U,43839
502
+ sglang/srt/models/minicpmo.py,sha256=b2TE13gXgfcBg9ibYaIHbFFtsMbPahs_qJC01BaMKu0,79521
503
+ sglang/srt/models/minicpmv.py,sha256=cMKLdprNl9siOlHJDY9OPRRDBCax84zHWCj16m-kzaA,40658
463
504
  sglang/srt/models/mistral.py,sha256=EYifJUUzN2Z2-iL37eJiNZF_DB0H4pa0mKlgYRIxM70,838
464
505
  sglang/srt/models/mixtral.py,sha256=6Fse2J-20IMylP-yzpEihIinaH37TmmslATbLcWBRYY,14926
465
506
  sglang/srt/models/mixtral_quant.py,sha256=MSa6UKPbgv8Rn8Iv8o1dQhcstAHLNQzE0eepFx_hYSw,15221
466
- sglang/srt/models/mllama.py,sha256=jG29hG_O6YzzakDP48hITQdMiNKX6fapkQ0duiyAD90,37887
507
+ sglang/srt/models/mllama.py,sha256=1hJiW6AHakH3QvGC2zKxuvZfo6YO5fbAmp5W-i2swB8,37891
467
508
  sglang/srt/models/olmo.py,sha256=FJk8A3T3TF5QcTV6rMP8np94QtvxpMWlgCsv_5VwpVE,12632
468
509
  sglang/srt/models/olmo2.py,sha256=U0ScFzWazOrb_Q90sfXkpVNAsXT-pgZbNgGh80R40VE,14288
469
510
  sglang/srt/models/olmoe.py,sha256=tx5OKWLOr6_pohe2eBcIodCmcuSjtpteHq_tG_QVYCY,15910
470
- sglang/srt/models/phi3_small.py,sha256=_ZDXVJN3B5f-46MMKxNim9874cVqJpzBipbmfBz6Hn4,15464
511
+ sglang/srt/models/phi3_small.py,sha256=6p-5EBbwN3FmhoL0VNrNb6VP7HqIhYgVBfQ98L6Rjjo,15469
471
512
  sglang/srt/models/qwen.py,sha256=edS0UYq6AoHZdYUJtQa5wyFNzZMW0JAMmBulH2uheaw,10719
472
- sglang/srt/models/qwen2.py,sha256=_jo9Ox7MrXCicw7Lhjla29UtRSXfsNd079So81rNcLM,15947
473
- sglang/srt/models/qwen2_5_vl.py,sha256=oA6Cz3-3RQN-D9B88s8dybV8idGjpdu2Fcm60i77Uqs,28184
513
+ sglang/srt/models/qwen2.py,sha256=D7VIFWbQ_MxY42K9j7ksgNTP9TNh_FA8u4FUND2xSuk,16038
514
+ sglang/srt/models/qwen2_5_vl.py,sha256=iZK_giWA58sb7cK9u_T-F3YgdgwDBq79nl2CszTL7Uo,24506
515
+ sglang/srt/models/qwen2_classification.py,sha256=dGrMm4ebd30_lBhHOhaV57ig2iOTx3nqB4GEzsrRIM8,2747
474
516
  sglang/srt/models/qwen2_eagle.py,sha256=Iz0HWL2FgSD3FqoFhfYmbIZeEYkPTJ96lYbkncmHJX4,4644
475
- sglang/srt/models/qwen2_moe.py,sha256=zYLJecN1mUyMBmnZoVaJd8LUKT4YZPBIO1lfFOqmU-c,17755
517
+ sglang/srt/models/qwen2_moe.py,sha256=9cLOPHBpwdID92Ed1CEjMUxMxcWX83lklunHbyK4To0,18206
476
518
  sglang/srt/models/qwen2_rm.py,sha256=-mQXDEv11p-I1HXgYLTtY6ROem6UYorO958WsDrzsgs,2837
477
- sglang/srt/models/qwen2_vl.py,sha256=G1Hx3n-5pXRtVHEtS-UjiQTr4AvltXwQ4PyroxNGfcE,23918
519
+ sglang/srt/models/qwen2_vl.py,sha256=JlcMUKKi2ynZR3vQN5gAwL13KEowq4Irc3Vj2WGpdnE,21983
478
520
  sglang/srt/models/registry.py,sha256=inKh9iwOp3LFYm3nqujg-OtABClOP-ifc1stA9cZegA,3434
479
521
  sglang/srt/models/stablelm.py,sha256=w93fNXpDwQbuKi4tdeo0bsXFZrMZVY4_pgNL0E5RErQ,12242
480
522
  sglang/srt/models/torch_native_llama.py,sha256=5tfFSMAXB3ScToqTALtCXa8Oo-qPCJh-KQCNB6QOlNA,19293
481
523
  sglang/srt/models/xverse.py,sha256=I7ivNsk6NRqPxlMUmdclpzDCvhAnWbv_GOj01MKHJrQ,13996
482
524
  sglang/srt/models/xverse_moe.py,sha256=xLwn5pRwQrvj7zMmwl3o49m7xILb2ACRdWvm9hY8LDc,16743
483
525
  sglang/srt/models/yivl.py,sha256=oToK7-u5IGO7xwpJIQ7VtudlK6-zPqJX4bt6_wv0SH8,4850
484
- sglang/srt/openai_api/adapter.py,sha256=gyzLXjV-FK393JOxGYzN2KIAPUXk9tfiSkI9Im0ikrE,67264
485
- sglang/srt/openai_api/protocol.py,sha256=6-vwYewe7JfmMMC56oRU13aRmmKUerZs50GPmiGeX7w,12490
526
+ sglang/srt/openai_api/adapter.py,sha256=wVy4N4hNve_N2empbLCDlAxC13jDFBAgCylU4ysYZzU,71334
527
+ sglang/srt/openai_api/protocol.py,sha256=MhGHrqywV7ghp9nsSLStl-lZN22RwPQTsSk3pUv-Nfs,13091
486
528
  sglang/srt/sampling/custom_logit_processor.py,sha256=tDvoLgLqn-sy1qcY6vSrpbnHCeqbdk0uhMOO-uy4p4E,1099
487
- sglang/srt/sampling/sampling_batch_info.py,sha256=T4UJ_CRB6A0HnRwEsqmxy1CJQMwZaVdxbdlCeTK4BUI,11992
488
- sglang/srt/sampling/sampling_params.py,sha256=HihGfhdR4FmOeltEqpW2kSLfNu94VCd8l0RNOQFSl-Q,5919
529
+ sglang/srt/sampling/sampling_batch_info.py,sha256=wrGGU27mWOi_yCBBCOvTQLBdyTjfkPuj7Hsk0zOFyH8,11989
530
+ sglang/srt/sampling/sampling_params.py,sha256=nXm44Inn91YtrMpAm5mDb6-97owRy-Bh6lZ0BIpw73I,5919
489
531
  sglang/srt/sampling/penaltylib/__init__.py,sha256=mtN8grFEcaBUhl4yBHmw8NNirt_i6uKO2cDNLHOpZQE,496
490
532
  sglang/srt/sampling/penaltylib/frequency_penalty.py,sha256=Loc3qjJTksNc5s-DV7QZHjgqoo5pxk7-nZzxwyhD2tQ,2144
491
533
  sglang/srt/sampling/penaltylib/min_new_tokens.py,sha256=rdU_D7RoIcrQPhysNQEzmr4TO2OoEi___p-i3QdwkgU,3331
492
534
  sglang/srt/sampling/penaltylib/orchestrator.py,sha256=XM-Lm6u7gYPtMZrTIc0FR4QxNZxBH5s_Cj82umyCzYk,5721
493
535
  sglang/srt/sampling/penaltylib/presence_penalty.py,sha256=NRh10AJrrQlGJ6S-enGdRefrTrWpyqrSm-aNnyqQNQQ,2119
494
- sglang/srt/speculative/build_eagle_tree.py,sha256=8_uUpkQAE0qcn5mA6NPjfl81EMuNxg4fZq628wjEZNU,20805
495
- sglang/srt/speculative/eagle_draft_cuda_graph_runner.py,sha256=zsF4AcsykaFjzZ5SKdvUQyEB1GVXTvR1_kI_oaQafBo,7704
496
- sglang/srt/speculative/eagle_utils.py,sha256=DhpNNNN56jw0ucXSBYq9_IzDuJmEKgXS6lbLczaghNs,27196
497
- sglang/srt/speculative/eagle_worker.py,sha256=ewqd3NxP0q8CHIhsigHYATv1SUnVl6SU7sUo5aLnFhE,20437
498
- sglang/srt/speculative/spec_info.py,sha256=hJR0b3gZ0QA0KZLq6TfqSDJkmpV6mThyle3sHYI4o4M,522
536
+ sglang/srt/speculative/build_eagle_tree.py,sha256=SFQ3eHbhfNxOdxgqDP5wSV_ZlIVqLw7VivycNZ963N0,11690
537
+ sglang/srt/speculative/eagle_draft_cuda_graph_runner.py,sha256=t2mbGZY23E7WraJJQW_4uaLTfnbPDExRVC7jldtuY6s,9287
538
+ sglang/srt/speculative/eagle_utils.py,sha256=_T3eRd59bzCR2YE_Z4W6Pux40KpcueWwN-rrLe-ikFQ,25812
539
+ sglang/srt/speculative/eagle_worker.py,sha256=OYHQPixTeFyGZ7MmKw6XWm62oTt2lb0LeoGq2e63G5I,24210
540
+ sglang/srt/speculative/spec_info.py,sha256=rhaKG0TzyF9XZYHEWp1jccwTBohSNsUDvxHFtAoOl18,709
541
+ sglang/test/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
499
542
  sglang/test/few_shot_gsm8k.py,sha256=7VLbWl4nCQs1wjtW4q-46jf9jUCycSs5Iw8v7sUSzBw,4284
500
543
  sglang/test/few_shot_gsm8k_engine.py,sha256=QQbrwOX6-cJDD3RZC_e7zPnt6aSo8JdF8X_lRHSjdDM,3886
501
544
  sglang/test/run_eval.py,sha256=9yO0hXZOcn4abEOs96T-XPguDEklK16Ltco0pGF3zCg,4020
502
- sglang/test/runners.py,sha256=4ZH0AtzMNFzxLXCxJlKpQOLl5C3jkut9YkDAvGYxW5w,29191
545
+ sglang/test/runners.py,sha256=l85tjnKQwBsZ3r1xwEluiHwUIsqNTDePBuSjPUo9TRQ,29305
503
546
  sglang/test/send_one.py,sha256=6FhbJ3c8RpXxvFTELRXaF97GpT7zXXsCDYZh1DqG22E,2550
504
547
  sglang/test/simple_eval_common.py,sha256=joqrGysuLnJFtzDRIgFkMsRyKUSyjVPFWp0_PHAL3Ik,12378
505
548
  sglang/test/simple_eval_gpqa.py,sha256=8Xt9Bw05c7SZTYrCZgB68OZUqUbLo69ywiyx0bTvSUk,3220
@@ -507,15 +550,18 @@ sglang/test/simple_eval_humaneval.py,sha256=zmV3xWYc2OrpiT9Dy55RTKZL5DEROD1cJ0NA
507
550
  sglang/test/simple_eval_math.py,sha256=6kGKNwNbLN-Af3Wj8WTimWhH-Xp3enDmSvvSjsgWUpk,2550
508
551
  sglang/test/simple_eval_mgsm.py,sha256=rd7TSUyxdKbrXaVoewo24V8lCo_6kO8zxPhhmvylpw8,10259
509
552
  sglang/test/simple_eval_mmlu.py,sha256=FkwamjGMjueTixymkedF-YiPloSLiy4ftILFUrKZ9XI,4357
510
- sglang/test/test_activation.py,sha256=jkdNRzJnbd5OgZliQaIXpxovlcky17UrweomcOcMxoE,1442
511
- sglang/test/test_block_fp8.py,sha256=bHYgQijsx0D4q0CgoZv6jOQLAWUz6QQC1cgT2oLLOnE,14420
512
- sglang/test/test_block_fp8_ep.py,sha256=hkuQjmCv3y_hWZj21cT9EaB6KSfT3JSzYPRQNFaLP-Q,10759
553
+ sglang/test/test_activation.py,sha256=GeTIJHxlLQfW3kM-X1FGa8Sa3dSGKHEXl5wEy-hfGis,1489
554
+ sglang/test/test_block_fp8.py,sha256=IqdQKt23annq_QR1gwVX0vzdMyWTEBLRhmPiLMemKI8,14458
555
+ sglang/test/test_block_fp8_ep.py,sha256=N1rvqbPErBaFFpeAw8TLYXGNZOoG7cfIBP2p5XbSyMo,10806
513
556
  sglang/test/test_custom_ops.py,sha256=XBTWh3jEsoe9hZ93p3LAUtjEj5l0qNEaZM5Mto6pbJA,3262
514
- sglang/test/test_layernorm.py,sha256=IacByD5d-stXjzBz8Ypamc7povlcedpKPbb_4JLgo3c,3720
557
+ sglang/test/test_dynamic_grad_mode.py,sha256=L76yUCuk_ymNpXD2CmO8r2GiGjIvD_gtTsuFDs2NolI,1638
558
+ sglang/test/test_layernorm.py,sha256=2GMWqqNDuGvSMSsEBF5eDCzwVSYA9E6hGhRo6s4ecKg,3764
515
559
  sglang/test/test_programs.py,sha256=VZ3vXtUDBnXz0M7gFdDH8hXg9Wa0j_qI8CVqjEgRN_E,18877
516
- sglang/test/test_utils.py,sha256=UBBk-F4A5hEtmRdu7vEttlnSmPVPJLQgtNZxMn1Tdo4,29188
517
- sglang-0.4.4.dist-info/LICENSE,sha256=FJXh51fvTQklojUFY89XVLsjxRcBqOxPs8XNy-2uZ0c,11346
518
- sglang-0.4.4.dist-info/METADATA,sha256=QgJKvztaH3uUY8TmfX6EPdgUpmR273E5K6PhTnww8B4,24305
519
- sglang-0.4.4.dist-info/WHEEL,sha256=52BFRY2Up02UkjOa29eZOS2VxUrpPORXg1pkohGGUS8,91
520
- sglang-0.4.4.dist-info/top_level.txt,sha256=yxhh3pYQkcnA7v3Bg889C2jZhvtJdEincysO7PEB09M,7
521
- sglang-0.4.4.dist-info/RECORD,,
560
+ sglang/test/test_utils.py,sha256=RbGOmiVOgDeKaTUS2KOLKtYHF3ZPQ8c2gp78Ddm7gF0,30685
561
+ sglang/test/attention/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
562
+ sglang/test/attention/test_flashattn_backend.py,sha256=OxS1KsPs19nwZcDtdURj7_liT1cIfEXb6W4FH9KMaaE,10808
563
+ sglang-0.4.4.post2.dist-info/licenses/LICENSE,sha256=FJXh51fvTQklojUFY89XVLsjxRcBqOxPs8XNy-2uZ0c,11346
564
+ sglang-0.4.4.post2.dist-info/METADATA,sha256=8OSdgzDzNxGk06NUSx9vVxQ-Nk2LnX-3ihUe2Sq2X40,24813
565
+ sglang-0.4.4.post2.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
566
+ sglang-0.4.4.post2.dist-info/top_level.txt,sha256=yxhh3pYQkcnA7v3Bg889C2jZhvtJdEincysO7PEB09M,7
567
+ sglang-0.4.4.post2.dist-info/RECORD,,
@@ -1,5 +1,5 @@
1
1
  Wheel-Version: 1.0
2
- Generator: setuptools (76.0.0)
2
+ Generator: setuptools (78.1.0)
3
3
  Root-Is-Purelib: true
4
4
  Tag: py3-none-any
5
5