sglang 0.4.4.post1__py3-none-any.whl → 0.4.4.post3__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 (185) 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 +26 -4
  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 +676 -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 +49 -8
  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/device_communicators/custom_all_reduce.py +1 -1
  34. sglang/srt/distributed/parallel_state.py +42 -8
  35. sglang/srt/entrypoints/engine.py +55 -5
  36. sglang/srt/entrypoints/http_server.py +78 -13
  37. sglang/srt/entrypoints/verl_engine.py +2 -0
  38. sglang/srt/function_call_parser.py +133 -55
  39. sglang/srt/hf_transformers_utils.py +28 -3
  40. sglang/srt/layers/activation.py +4 -2
  41. sglang/srt/layers/attention/base_attn_backend.py +1 -1
  42. sglang/srt/layers/attention/flashattention_backend.py +434 -0
  43. sglang/srt/layers/attention/flashinfer_backend.py +1 -1
  44. sglang/srt/layers/attention/flashmla_backend.py +284 -0
  45. sglang/srt/layers/attention/triton_backend.py +171 -38
  46. sglang/srt/layers/attention/triton_ops/decode_attention.py +94 -31
  47. sglang/srt/layers/attention/triton_ops/extend_attention.py +14 -5
  48. sglang/srt/layers/attention/utils.py +53 -0
  49. sglang/srt/layers/attention/vision.py +9 -28
  50. sglang/srt/layers/dp_attention.py +41 -19
  51. sglang/srt/layers/layernorm.py +24 -2
  52. sglang/srt/layers/linear.py +17 -5
  53. sglang/srt/layers/logits_processor.py +25 -7
  54. sglang/srt/layers/moe/ep_moe/kernels.py +110 -11
  55. sglang/srt/layers/moe/ep_moe/layer.py +273 -1
  56. sglang/srt/layers/moe/ep_moe/token_dispatcher.py +416 -0
  57. sglang/srt/layers/moe/fused_moe_native.py +2 -1
  58. sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=64,device_name=NVIDIA_L20,dtype=int8_w8a8.json +146 -0
  59. sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=64,device_name=NVIDIA_L40S,dtype=int8_w8a8.json +146 -0
  60. 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
  61. 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
  62. sglang/srt/layers/moe/fused_moe_triton/fused_moe.py +23 -32
  63. sglang/srt/layers/moe/fused_moe_triton/layer.py +1 -2
  64. sglang/srt/layers/moe/topk.py +60 -20
  65. sglang/srt/layers/parameter.py +1 -1
  66. sglang/srt/layers/quantization/__init__.py +80 -53
  67. sglang/srt/layers/quantization/awq.py +200 -0
  68. sglang/srt/layers/quantization/base_config.py +5 -0
  69. sglang/srt/layers/quantization/blockwise_int8.py +1 -1
  70. sglang/srt/layers/quantization/compressed_tensors/__init__.py +0 -0
  71. sglang/srt/layers/quantization/compressed_tensors/compressed_tensors.py +652 -0
  72. sglang/srt/layers/quantization/compressed_tensors/compressed_tensors_moe.py +658 -0
  73. sglang/srt/layers/quantization/compressed_tensors/schemes/__init__.py +9 -0
  74. sglang/srt/layers/quantization/compressed_tensors/schemes/compressed_tensors_scheme.py +56 -0
  75. sglang/srt/layers/quantization/compressed_tensors/schemes/compressed_tensors_w8a8_fp8.py +162 -0
  76. sglang/srt/layers/quantization/compressed_tensors/utils.py +218 -0
  77. sglang/srt/layers/quantization/fp8.py +76 -34
  78. sglang/srt/layers/quantization/fp8_kernel.py +25 -8
  79. sglang/srt/layers/quantization/fp8_utils.py +284 -28
  80. sglang/srt/layers/quantization/gptq.py +36 -19
  81. sglang/srt/layers/quantization/kv_cache.py +98 -0
  82. sglang/srt/layers/quantization/modelopt_quant.py +9 -7
  83. sglang/srt/layers/quantization/utils.py +153 -0
  84. sglang/srt/layers/quantization/w8a8_fp8.py +70 -19
  85. sglang/srt/layers/rotary_embedding.py +78 -87
  86. sglang/srt/layers/sampler.py +1 -1
  87. sglang/srt/lora/backend/base_backend.py +4 -4
  88. sglang/srt/lora/backend/flashinfer_backend.py +12 -9
  89. sglang/srt/lora/backend/triton_backend.py +5 -8
  90. sglang/srt/lora/layers.py +87 -33
  91. sglang/srt/lora/lora.py +2 -22
  92. sglang/srt/lora/lora_manager.py +67 -30
  93. sglang/srt/lora/mem_pool.py +117 -52
  94. sglang/srt/lora/triton_ops/gate_up_lora_b.py +10 -4
  95. sglang/srt/lora/triton_ops/qkv_lora_b.py +8 -3
  96. sglang/srt/lora/triton_ops/sgemm_lora_a.py +16 -5
  97. sglang/srt/lora/triton_ops/sgemm_lora_b.py +11 -6
  98. sglang/srt/lora/utils.py +18 -1
  99. sglang/srt/managers/cache_controller.py +2 -5
  100. sglang/srt/managers/data_parallel_controller.py +30 -8
  101. sglang/srt/managers/expert_distribution.py +81 -0
  102. sglang/srt/managers/io_struct.py +43 -5
  103. sglang/srt/managers/mm_utils.py +373 -0
  104. sglang/srt/managers/multimodal_processor.py +68 -0
  105. sglang/srt/managers/multimodal_processors/base_processor.py +275 -0
  106. sglang/srt/managers/multimodal_processors/clip.py +63 -0
  107. sglang/srt/managers/multimodal_processors/deepseek_vl_v2.py +119 -0
  108. sglang/srt/managers/multimodal_processors/gemma3.py +83 -0
  109. sglang/srt/managers/{image_processors → multimodal_processors}/janus_pro.py +20 -15
  110. sglang/srt/managers/{image_processors → multimodal_processors}/llava.py +10 -15
  111. sglang/srt/managers/multimodal_processors/minicpm.py +167 -0
  112. sglang/srt/managers/{image_processors → multimodal_processors}/mlama.py +7 -8
  113. sglang/srt/managers/{image_processors → multimodal_processors}/qwen_vl.py +28 -22
  114. sglang/srt/managers/schedule_batch.py +134 -30
  115. sglang/srt/managers/scheduler.py +290 -31
  116. sglang/srt/managers/session_controller.py +1 -1
  117. sglang/srt/managers/tokenizer_manager.py +59 -24
  118. sglang/srt/managers/tp_worker.py +4 -1
  119. sglang/srt/managers/tp_worker_overlap_thread.py +3 -3
  120. sglang/srt/managers/utils.py +6 -1
  121. sglang/srt/mem_cache/hiradix_cache.py +18 -7
  122. sglang/srt/mem_cache/memory_pool.py +255 -98
  123. sglang/srt/mem_cache/paged_allocator.py +2 -2
  124. sglang/srt/mem_cache/radix_cache.py +4 -4
  125. sglang/srt/model_executor/cuda_graph_runner.py +36 -21
  126. sglang/srt/model_executor/forward_batch_info.py +68 -11
  127. sglang/srt/model_executor/model_runner.py +75 -8
  128. sglang/srt/model_loader/loader.py +171 -3
  129. sglang/srt/model_loader/weight_utils.py +51 -3
  130. sglang/srt/models/clip.py +563 -0
  131. sglang/srt/models/deepseek_janus_pro.py +31 -88
  132. sglang/srt/models/deepseek_nextn.py +22 -10
  133. sglang/srt/models/deepseek_v2.py +329 -73
  134. sglang/srt/models/deepseek_vl2.py +358 -0
  135. sglang/srt/models/gemma3_causal.py +694 -0
  136. sglang/srt/models/gemma3_mm.py +468 -0
  137. sglang/srt/models/llama.py +47 -7
  138. sglang/srt/models/llama_eagle.py +1 -0
  139. sglang/srt/models/llama_eagle3.py +196 -0
  140. sglang/srt/models/llava.py +3 -3
  141. sglang/srt/models/llavavid.py +3 -3
  142. sglang/srt/models/minicpmo.py +1995 -0
  143. sglang/srt/models/minicpmv.py +62 -137
  144. sglang/srt/models/mllama.py +4 -4
  145. sglang/srt/models/phi3_small.py +1 -1
  146. sglang/srt/models/qwen2.py +3 -0
  147. sglang/srt/models/qwen2_5_vl.py +68 -146
  148. sglang/srt/models/qwen2_classification.py +75 -0
  149. sglang/srt/models/qwen2_moe.py +9 -1
  150. sglang/srt/models/qwen2_vl.py +25 -63
  151. sglang/srt/openai_api/adapter.py +201 -104
  152. sglang/srt/openai_api/protocol.py +33 -7
  153. sglang/srt/patch_torch.py +71 -0
  154. sglang/srt/sampling/sampling_batch_info.py +1 -1
  155. sglang/srt/sampling/sampling_params.py +6 -6
  156. sglang/srt/server_args.py +114 -14
  157. sglang/srt/speculative/build_eagle_tree.py +7 -347
  158. sglang/srt/speculative/eagle_draft_cuda_graph_runner.py +41 -5
  159. sglang/srt/speculative/eagle_utils.py +208 -252
  160. sglang/srt/speculative/eagle_worker.py +140 -54
  161. sglang/srt/speculative/spec_info.py +6 -1
  162. sglang/srt/torch_memory_saver_adapter.py +22 -0
  163. sglang/srt/utils.py +215 -21
  164. sglang/test/__init__.py +0 -0
  165. sglang/test/attention/__init__.py +0 -0
  166. sglang/test/attention/test_flashattn_backend.py +312 -0
  167. sglang/test/runners.py +29 -2
  168. sglang/test/test_activation.py +2 -1
  169. sglang/test/test_block_fp8.py +5 -4
  170. sglang/test/test_block_fp8_ep.py +2 -1
  171. sglang/test/test_dynamic_grad_mode.py +58 -0
  172. sglang/test/test_layernorm.py +3 -2
  173. sglang/test/test_utils.py +56 -5
  174. sglang/utils.py +31 -0
  175. sglang/version.py +1 -1
  176. {sglang-0.4.4.post1.dist-info → sglang-0.4.4.post3.dist-info}/METADATA +16 -8
  177. {sglang-0.4.4.post1.dist-info → sglang-0.4.4.post3.dist-info}/RECORD +180 -132
  178. {sglang-0.4.4.post1.dist-info → sglang-0.4.4.post3.dist-info}/WHEEL +1 -1
  179. sglang/srt/configs/qwen2_5_vl_config.py +0 -1006
  180. sglang/srt/managers/image_processor.py +0 -55
  181. sglang/srt/managers/image_processors/base_image_processor.py +0 -219
  182. sglang/srt/managers/image_processors/minicpmv.py +0 -86
  183. sglang/srt/managers/multi_modality_padding.py +0 -134
  184. {sglang-0.4.4.post1.dist-info → sglang-0.4.4.post3.dist-info/licenses}/LICENSE +0 -0
  185. {sglang-0.4.4.post1.dist-info → sglang-0.4.4.post3.dist-info}/top_level.txt +0 -0
@@ -1,107 +1,126 @@
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=9w5bzcQlm9bo7vWTq3XUJIaKvZy5k2vl4CNjgGq2LV8,56095
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=4ZEzfisb7CBE8mZJHsVVkqSmecAalLdtwfQHDTnpWTw,28
11
+ sglang/utils.py,sha256=Ns1-swsHRb8jYA0wWCQjCR_CDBWIJgyoooEng4fsqQc,16155
12
+ sglang/version.py,sha256=Lhj8wFyI_a4P5gAvUzUTThFvAemjjMQEp_O1QuZa64c,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=k_L4Ex4Lc5TUNWiCO6BcSJjlv2TQmvphyGZIxHfl0e0,20835
32
- sglang/srt/hf_transformers_utils.py,sha256=FPoOvGdEy-lwrOOqHaQnb_tJCGmCW3CpKUjleY7Cwj4,8221
32
+ sglang/srt/function_call_parser.py,sha256=buYENeNEP5bhsvD424yGCa9wOqSfVOZSRn6zLiSJp5I,23733
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
36
+ sglang/srt/patch_torch.py,sha256=Fw_QGqSsAdyCumi0dT2HyPlppf9xd3-tQPvwuBZfhxU,2625
35
37
  sglang/srt/reasoning_parser.py,sha256=45xsU9RCPfyG4_Zx4y3-JPyNgAtrqwKI4j5R2NT4g1s,5594
36
38
  sglang/srt/server.py,sha256=PrQb9r6L9syWHKlggbbiQYsKtpwSmECqozRbf8qnoV8,874
37
- sglang/srt/server_args.py,sha256=ZKYb3tL6m8pVNH-3m9yRRRMkz2mlLkRhV0acME7dn0Q,44864
38
- sglang/srt/torch_memory_saver_adapter.py,sha256=--FgbrcvJxTcRe856plD9ktqgrHGPTE18eZCJlE50hY,1255
39
- sglang/srt/utils.py,sha256=Nh85Igdbmi5nN09s2rFsnnAIUcijk5jljmktiYByZp8,50163
39
+ sglang/srt/server_args.py,sha256=SEZjYkfjwv0V1xQafh7Zex7NS5MVenfKWytcd4kZ3t8,49024
40
+ sglang/srt/torch_memory_saver_adapter.py,sha256=KG3wM9-xZsSdsmORofArnNR7hH55GEyFxaderCDcK9w,1853
41
+ sglang/srt/utils.py,sha256=qoo6k_-uxthkMhotF9ba_-njOV-v1AbEcJ9XdQwcha8,56116
40
42
  sglang/srt/warmup.py,sha256=FmJiYfjRr3X_eAe7ojQaPoN17LvHpjDmRWRnO-k86AQ,1469
41
- sglang/srt/configs/__init__.py,sha256=iDfWab2iDMQ3F5bezHjM-fHi7u8iKqk_UBLZFabXxiI,470
43
+ sglang/srt/configs/__init__.py,sha256=vulncVn70WqIT6s0HaB8p_Q6FjOiaLwNZWpoJS9FIuQ,399
42
44
  sglang/srt/configs/chatglm.py,sha256=j-b0YkdYUmQm2y1kNmMJtKeACxWKmBbvNNkDWbs6kbI,2907
43
45
  sglang/srt/configs/dbrx.py,sha256=tdhIkXAQl1yr0MxqFmsDG1E0e2puRTTKm6UTyANBLac,11005
46
+ sglang/srt/configs/deepseekvl2.py,sha256=qeI7TO4f2DHIxCRiuRxpnP17K5lCMSNGJjEcq2yF9Ds,23328
44
47
  sglang/srt/configs/device_config.py,sha256=kfmpPOECqYxcRoY-ko0QZRhyiBWUGP2CMF51DMUN5nU,435
45
48
  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
49
+ sglang/srt/configs/janus_pro.py,sha256=-QtJ4ZGZiAJb0AkOEcuCHzIKLw23nF8nRk3rdCcoUO0,19016
50
+ sglang/srt/configs/load_config.py,sha256=qs-AxuplouBx2tsv9KGBOLZPbwzuVA4vbktbGP_cRp8,3309
51
+ sglang/srt/configs/model_config.py,sha256=U8cynUgfijwutxZ-8hhoaWTKR5NqpXGIrGxlhbWSBlY,20556
52
+ sglang/srt/configs/utils.py,sha256=3nHUfisMs_Ltuhv8OZTNCJp63YJKJVF43h1QZB1zqx8,670
53
+ sglang/srt/connector/__init__.py,sha256=czLX5JOxuMhH-T9eSJzoc1qv1B4z9chyffDRL5I6wo4,1247
54
+ sglang/srt/connector/base_connector.py,sha256=i6i1TIzsz4NbSEkrdMPq-urb2sN2aLAx8dazga4gB9U,2833
55
+ sglang/srt/connector/redis.py,sha256=KoLLUSY4Lu9zUrRJy2yYdtf1yanTaS0bQP8V0t95DB0,2552
56
+ sglang/srt/connector/s3.py,sha256=hn5Ow9Y6ItS8RMJDwnTQe5QaiQ7HKMgiwggE9QnUewM,3743
57
+ sglang/srt/connector/utils.py,sha256=isTvxauz1_8no5MW6p8Bwj2H9mQkweaRO_RSwAPA7R8,983
58
+ sglang/srt/connector/serde/__init__.py,sha256=MvBJ7FBJtnou-AD-EdsCdAIDCcA8kWvUNuRViILVJ-0,718
59
+ sglang/srt/connector/serde/safe_serde.py,sha256=Iv_mFsDvk-cXuw0WTykBZ2yGxW0jb82DwVqC7r3RmAU,750
60
+ sglang/srt/connector/serde/serde.py,sha256=n59I2MXLa7WCyN_8pEd8L-scJk7lMhmEX-GOUIhF0ZA,1004
50
61
  sglang/srt/constrained/base_grammar_backend.py,sha256=q2RTH9hv2nKUF_-HVdZzEUjG6LxejPqXjvUOE1NDWaU,6788
51
62
  sglang/srt/constrained/llguidance_backend.py,sha256=ej7wN13SzCsT310C6OIyUg2zs5jeuLl3Ocok9SP9-c4,5702
52
63
  sglang/srt/constrained/outlines_backend.py,sha256=UWv2xjg8x4XtoqpY8LoorlJaYOZhfDeIr5YCiFn4knA,6812
53
64
  sglang/srt/constrained/outlines_jump_forward.py,sha256=iZWXeR3gNYoMubLGyFmLPO4V2YsN5DiGjD71Xk9iFaE,6418
54
65
  sglang/srt/constrained/xgrammar_backend.py,sha256=W7_qyyQiOUwejIPCnWgJrp6ka5fy137SiJtxt3VNruM,6220
66
+ sglang/srt/disaggregation/conn.py,sha256=amOujTy2jFwdfYhGxuTLAMNWGPbIKGoAwWDqKxZ06gc,1950
67
+ sglang/srt/disaggregation/decode.py,sha256=5pgXeIQBBJXQpVXpRm7vAauRmy5DtIi8953dZFBAPeA,18075
68
+ sglang/srt/disaggregation/mini_lb.py,sha256=upwG_4CdurUUiPuYGUO4OJQu22lDx6gnsM0xKv5QRmg,10692
69
+ sglang/srt/disaggregation/prefill.py,sha256=zw8hDy6Txq_MpC5j0fndLNcKoypT2BhxTkqqTuilMCE,9053
70
+ sglang/srt/disaggregation/utils.py,sha256=ebOZ3lSFVkbNtl6uUfS6sYYYVBjgmWdQLOsqIZBGgN0,1088
55
71
  sglang/srt/distributed/__init__.py,sha256=jFOcyt-wFAPMBUAf9zkZalNQlt-4rqmT6pCKBz1E4qo,149
56
72
  sglang/srt/distributed/communication_op.py,sha256=IBnFUdMftK_VSTMMMitGveonorFUUVNL4guqO31cMSc,1130
57
- sglang/srt/distributed/parallel_state.py,sha256=lgsscyYrqJLvblF6c5JdfYQrT-YfOKBPCHZnhazlD28,49614
73
+ sglang/srt/distributed/parallel_state.py,sha256=hoTgLYfHIKMb_tSwBTauuusJZ8oY9BsiubTTOF8UfIw,50713
58
74
  sglang/srt/distributed/utils.py,sha256=U-BSaXYjWwnfG8g-tUfBhjKt5Ug097nyHtu3g3aea_Y,8473
59
75
  sglang/srt/distributed/device_communicators/cuda_wrapper.py,sha256=3jvPG-Ow5UBLiXhfx8T8snR7crSZbPpARAggsDPWq7k,7038
60
- sglang/srt/distributed/device_communicators/custom_all_reduce.py,sha256=kmbAV_W9ObjI7NH8X0gV__D9iHCuj6QDyhU5ZKhEQD8,22209
76
+ sglang/srt/distributed/device_communicators/custom_all_reduce.py,sha256=iLCrFQ3dyH_xZL3vI-paTpxeP7Rt4DszRIk8qdbtfvA,22214
61
77
  sglang/srt/distributed/device_communicators/custom_all_reduce_utils.py,sha256=q2q1A_Sqvrvkrgf7Tjg5XhXR1JWzzUUPHSicAKK2SjE,11022
62
78
  sglang/srt/distributed/device_communicators/hpu_communicator.py,sha256=gPjEH1-izoby5uDrfUlzNf21luPT0Ow7pJjhCRKnHy8,1728
63
79
  sglang/srt/distributed/device_communicators/pynccl.py,sha256=G-Dut_QJHOUG0j7--ZqapHtvm70Lgl7obtE6ZfgeAiU,10064
64
80
  sglang/srt/distributed/device_communicators/pynccl_wrapper.py,sha256=LblisImY9d6EMz-oPS9J16WHo2Q_SRL1DtlJKK63Hfg,15349
65
81
  sglang/srt/distributed/device_communicators/shm_broadcast.py,sha256=bbruDIM1GgKIdB6gi71_I0mpB179I-qyvwKuSj1Kaic,20816
66
82
  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
69
- 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=gaWG5c6Np2OJls4ldZ5Ea2etpY3T-0uyru7IKBuEtVM,7199
83
+ sglang/srt/entrypoints/engine.py,sha256=b4JmlvK4M2cWfcbc5NOlR7HjkNzDENlzL4rbSaPjBdU,21295
84
+ sglang/srt/entrypoints/http_server.py,sha256=VM-gVwW-Ef_SikxoVSE06yydlLo6TGJykjKgffnGoXY,27104
85
+ sglang/srt/entrypoints/verl_engine.py,sha256=PypBCkUJYy7iP3mKB-W0KYkjZzs4Rq6DqxNgG_nBZaM,5903
86
+ sglang/srt/layers/activation.py,sha256=1ykXZO0BGz7DFVE-EK26b02I5AgH2IuU4PQB6oUcF4M,6003
87
+ sglang/srt/layers/dp_attention.py,sha256=fC1kaYkHGoFjZ2KHTzPFW0e93El-XLRt7ZidkXYIVhE,7595
72
88
  sglang/srt/layers/elementwise.py,sha256=y2mQqjbF2FmFtNYBk5ecTyaj3ELoZyz-rWPY8rrxCtk,13765
73
- sglang/srt/layers/layernorm.py,sha256=DI2Ih34bLh5Ex_yd8-X8L7kNRcYStpc9XpIS4zIoM_U,3860
74
- sglang/srt/layers/linear.py,sha256=8HyDjnpUu1ZxdqVRt9jiuO4GvTr2JsO42jwgtl10Qjs,51063
75
- sglang/srt/layers/logits_processor.py,sha256=RHvuPECoMpdU4Fpv1UrQ3yVz9qHKfbYxLS6k6bNRh0s,23243
76
- sglang/srt/layers/parameter.py,sha256=-jaiI-dIhiSPigwcQGydBiKEMSVg7XuMcg8Bzb8BHMo,15162
89
+ sglang/srt/layers/layernorm.py,sha256=189bORMggKhYcEYEjl6JRcuIoUPllHo3SheoH6YiORY,4546
90
+ sglang/srt/layers/linear.py,sha256=HYIGxpRYL6x-jNOkyNtGAw5Ak9Nq8jkntddgTBER_1w,51486
91
+ sglang/srt/layers/logits_processor.py,sha256=Vp8ibljVEezTr54xzeOcjiJR7JdYO8ItkO5nLIIMVu0,24206
92
+ sglang/srt/layers/parameter.py,sha256=0OTMtmsNds42e3z3wHTRJiUfxCWFwSL6DHrqgeTgGt8,15151
77
93
  sglang/srt/layers/pooler.py,sha256=rj2lygvleBnyLCBZ8I11HGMgpfIDsT0l3PIkshJwdu4,1606
78
94
  sglang/srt/layers/radix_attention.py,sha256=UDL0y4Zasay_Rk-_XmIU4kaGbaF26ONvEHX5EQzLrqI,2260
79
- sglang/srt/layers/rotary_embedding.py,sha256=3DVmZ9UUYXPiWVxE9Ay25S0g-Oeh-_u7PXvPkRJ8nnA,44032
80
- sglang/srt/layers/sampler.py,sha256=NIHFR9oot9q-4n414zS3K3oLAHXtuT27mfv-3KgH7Dc,11638
95
+ sglang/srt/layers/rotary_embedding.py,sha256=W4w6Rn-7AgdOqrgVxWY8QJwSpiKUrRbH03bmX6hdw1I,43175
96
+ sglang/srt/layers/sampler.py,sha256=yipSyN5UWGwGS-BC-WzWMmelys4CCDtK_8b1OpaK6sM,11622
81
97
  sglang/srt/layers/torchao_utils.py,sha256=Ws24FdRBSkTpyeyA6bQrdDm-W5wfDxKvSIPUSahyMfA,4063
82
98
  sglang/srt/layers/vocab_parallel_embedding.py,sha256=QUxd4sELx6p3dHvEKmccPZ-phdd_9EjNdwjH3SJ9zxI,22238
83
- sglang/srt/layers/attention/base_attn_backend.py,sha256=j6pDIDuxlWVmc8GzNxcAFi1USRo4gqqI7G65XAQ5u5M,3263
99
+ sglang/srt/layers/attention/base_attn_backend.py,sha256=X_GIbQuU9njtUEGdUP7E_KRhmGxj3UyPHNESlL3QaQ8,3264
84
100
  sglang/srt/layers/attention/double_sparsity_backend.py,sha256=2ZRL_gYz14idoVqQzeQ6N77nXer0f_8_TUYw40XUUz0,9161
85
- sglang/srt/layers/attention/flashinfer_backend.py,sha256=iSJ7wSlKWB2GmumG1WtakhPOjQmcI60wz5ZBeiayIHU,45980
101
+ sglang/srt/layers/attention/flashattention_backend.py,sha256=kiiuklwtEKCgKjQPgvNnl9zjX8IGYqti4j21b-DcRGM,16831
102
+ sglang/srt/layers/attention/flashinfer_backend.py,sha256=m1rOYGr9uaxeUPA8BGPGGe8yicLJKNmRm4zxkJoQU9k,45980
86
103
  sglang/srt/layers/attention/flashinfer_mla_backend.py,sha256=BgbGK5iROulOco7o3vYTaPw1Ei3EaQj6Cn9W57Z4amQ,30360
104
+ sglang/srt/layers/attention/flashmla_backend.py,sha256=iWzxEEIJ6g3RrovSReqAZMjjcv36qCIsPEHZGs_J8JI,10543
87
105
  sglang/srt/layers/attention/torch_native_backend.py,sha256=KABmBrMqKa4x08kkQYdIcZUGydvmaVJIUfo3y8jhFHI,9270
88
- sglang/srt/layers/attention/triton_backend.py,sha256=CiDI97Yp1mPOpeObpr2F8z-0OU8EPf4mQ9sQzK-8lJE,21520
89
- sglang/srt/layers/attention/utils.py,sha256=gwZoJDyJ9OIPMHWl6r3qkuyVp4Sji9juX7Pwvh9PNxI,1131
90
- sglang/srt/layers/attention/vision.py,sha256=4DjJfK4aUF994R2c_R1HQ6Q7R_CrWIyBT9NGR0yQsvY,12471
91
- sglang/srt/layers/attention/triton_ops/decode_attention.py,sha256=OcN_uL6dwAU3ImCqDHH3ZamJCQSofI8C7erTBc-ZPps,17926
106
+ sglang/srt/layers/attention/triton_backend.py,sha256=cyxOaUU1CNhaEezJH9j0dd20cwxwIVGGN3jNXFTVkIY,26714
107
+ sglang/srt/layers/attention/utils.py,sha256=J9mA-cbZT3uTlaKXo0HEAaeMei_TS2o4McTna9LVDCE,2750
108
+ sglang/srt/layers/attention/vision.py,sha256=hUUkMyhl2WbKAY8ykvolBWgIMHuihiZ7w_UkJVxUQUo,11890
109
+ sglang/srt/layers/attention/triton_ops/decode_attention.py,sha256=DPu_aCPgwPqKWZPEQmp_xA7MPbpV2ip-MEICCB470Ao,19120
92
110
  sglang/srt/layers/attention/triton_ops/double_sparsity_attention.py,sha256=BXUY8ARHBF2s9x9waiEwfZwcMgvuaJA0gxb4OeUZ_tY,31167
93
- sglang/srt/layers/attention/triton_ops/extend_attention.py,sha256=SZezMpXMr2AEEIvWBQYzx_Gnmr2fyFhQET8eucdmfdU,13244
111
+ sglang/srt/layers/attention/triton_ops/extend_attention.py,sha256=hbBvnhc2zqu-E3HNROVXyNOZbtDkVRuFus-yTjmE0Sg,13668
94
112
  sglang/srt/layers/attention/triton_ops/prefill_attention.py,sha256=Y66gZ37u0GKMPtI8n5MbO6uOxRuGEmKIG0IPbJTOqAM,6213
95
113
  sglang/srt/layers/attention/triton_ops/rocm_mla_decode_rope.py,sha256=664WnAJ91EiCUZOcnVDfbTQf4uGJ4ZDZB1CbxpEUFZc,13866
96
- sglang/srt/layers/moe/fused_moe_native.py,sha256=domK1jDc5-zJcdwDso-YgBp_O136VHifnvFVlfFvKsY,4325
114
+ sglang/srt/layers/moe/fused_moe_native.py,sha256=KF0J5W5biWYWbERUHAduQFnUV5om9VaYv0I9avsVfgU,4330
97
115
  sglang/srt/layers/moe/router.py,sha256=gvyK7hXlujfCZCmAIFc3oxfgjuAjzlpPe3mp1Blc6Y0,10419
98
- sglang/srt/layers/moe/topk.py,sha256=MOwBNBwxxW__mgmpX08RIqrh77aXLEgr0F5b5iF7hRI,7512
116
+ sglang/srt/layers/moe/topk.py,sha256=udkEhgzXbNqk0-bSReeLMDtP_3aOtTlZa6qPn82keVs,8436
99
117
  sglang/srt/layers/moe/ep_moe/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
100
- sglang/srt/layers/moe/ep_moe/kernels.py,sha256=hG7E29bbj8HeXEkAd6N0HzJZPn7cwVx1VJptzpRWOv0,15376
101
- sglang/srt/layers/moe/ep_moe/layer.py,sha256=y8YaVIvzZ5bomc_9wvc1SwcgOddmVPlD5FYQfxI1kPY,27890
118
+ sglang/srt/layers/moe/ep_moe/kernels.py,sha256=E-VyToHN13f443U3UboDgLwv-t8gAcXOtPHkrI2xUpI,18793
119
+ sglang/srt/layers/moe/ep_moe/layer.py,sha256=ZiS9viPdcpB7SjylNVB3DSuF3Az1jGjDPdCqd657Qf0,36990
120
+ sglang/srt/layers/moe/ep_moe/token_dispatcher.py,sha256=SPHfAXIkZmFqXMU0a2eob9j_cni269v5jevGm2x1vWM,14845
102
121
  sglang/srt/layers/moe/fused_moe_triton/__init__.py,sha256=h9yMFAL_bagUf-qBED8gSWdCOb7d8IdA-pE-L_nIg8E,842
103
- sglang/srt/layers/moe/fused_moe_triton/fused_moe.py,sha256=R1XYyB-oS0QMxIoGrMKPBlk17Jq-z1dTzafZ7bcTSwU,41856
104
- sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=tHSxJ2JWwaSwPuyleILJiMH4cNrjf9P5Y8dcAJ8gm5o,24129
122
+ sglang/srt/layers/moe/fused_moe_triton/fused_moe.py,sha256=lBF40JXOgUzFMza020gRYol4WdLME9OB-YA9tSfsJ3k,41516
123
+ sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=CYT3ujh5ifonhqQc1uYSa6maJADUOBJoRHMIWmf-DU4,24100
105
124
  "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
106
125
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=1,N=14336,device_name=NVIDIA_A100-SXM4-80GB.json",sha256=JJN0hryyLr5Zv3dSS7C8cPFhAwTT6XxUVnBGMZvV6JA,2752
107
126
  "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
@@ -146,6 +165,9 @@ sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=tHSxJ2JWwaSwPuyleILJiMH4c
146
165
  "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
147
166
  "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
148
167
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=64,device_name=NVIDIA_A800-SXM4-80GB.json",sha256=Y4WpmVGlGERHoeoQNGkQ-GC-MsEtMblqnAVuDbARJdw,3240
168
+ "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
169
+ "sglang/srt/layers/moe/fused_moe_triton/configs/E=256,N=64,device_name=NVIDIA_L40S,dtype=int8_w8a8.json",sha256=Sc9xK1wtRUqIzXppbutcq-Y2e9M0DZl2OGVzzB0aQuI,3265
170
+ "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
149
171
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=1280,device_name=NVIDIA_A100-SXM4-80GB.json",sha256=6QPLhZurIqcMVdy3w0Dd7gLViKxsyJRBz-qd8agpi6Q,3248
150
172
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=1280,device_name=NVIDIA_A800-SXM4-80GB.json",sha256=uv-RwTNZT2n264dLo4eWxUpB3g7QqUyf2MFEGiRvoqQ,3251
151
173
  "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
@@ -159,6 +181,7 @@ sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=tHSxJ2JWwaSwPuyleILJiMH4c
159
181
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=320,device_name=NVIDIA_H100_80GB_HBM3.json",sha256=prj-QxdfS9Ns5WRPvahY_Tr7CyqlaVgNHPT89SS5zzg,3239
160
182
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=320,device_name=NVIDIA_H200,dtype=fp8_w8a8.json",sha256=AUgoHK1PmAFehSNmsbxunlBdzM50Q5nFvdnG9FSOjOw,3265
161
183
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=320,device_name=NVIDIA_H200.json",sha256=pC5fdtEFc5aVNzpj_REHhz1QPrGvgI9iQCvlodDP7J8,3244
184
+ "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
162
185
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=640,device_name=NVIDIA_A100-SXM4-80GB.json",sha256=oxOKFDrgmw1YmgxTtRa1uoe3p09ylTLrkj_jOTqNh1Q,3249
163
186
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=64,N=640,device_name=NVIDIA_A800-SXM4-80GB.json",sha256=bfr70r6PmM95w7raabQOaOOSPiwU2OQCOZh-kKXIehY,3248
164
187
  "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
@@ -217,18 +240,28 @@ sglang/srt/layers/moe/fused_moe_triton/layer.py,sha256=tHSxJ2JWwaSwPuyleILJiMH4c
217
240
  "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
218
241
  "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
219
242
  "sglang/srt/layers/moe/fused_moe_triton/configs/E=8,N=8192,device_name=NVIDIA_H200,dtype=fp8_w8a8.json",sha256=Uz5X80VcNBOaxshwVNUEittHk2zqB4HQCfTJ4TPG5aM,3274
220
- sglang/srt/layers/quantization/__init__.py,sha256=eBbrWZLYhjtzva61n0VTvs4Q60hb2S8qlHRuiPSFiME,11086
221
- sglang/srt/layers/quantization/base_config.py,sha256=daK9p0aijMszLUm1W4Pc33FK87MdqYK1NoWFKif-j80,4599
222
- sglang/srt/layers/quantization/blockwise_int8.py,sha256=AhMwTDFkE6NmxO_ER6rq1K3k5Bo5ZZAFk1OmEju7rEo,14908
223
- sglang/srt/layers/quantization/fp8.py,sha256=pX94MLbzeDuwSgagKux8MInW4oeZ0YEm_j4A6mDq62o,39758
224
- sglang/srt/layers/quantization/fp8_kernel.py,sha256=BNQC6MlDGw5HnaSBfmn7vajoLEE57J8boAy_2KmJ7DU,23816
225
- sglang/srt/layers/quantization/fp8_utils.py,sha256=IOhwQv7eDT0-L2H9gGOPx5gu7lGJxUHDubltAnNU5ps,11100
226
- sglang/srt/layers/quantization/gptq.py,sha256=u-WadiLnVbeskCU8Ar-1IYofraOf1caYkGrUFFYC27k,14651
243
+ sglang/srt/layers/quantization/__init__.py,sha256=ppIxDMFk-XcmDS7SWZe5fc9ymJIu3H3rCHTqm6utKCU,12001
244
+ sglang/srt/layers/quantization/awq.py,sha256=VImnVCU_QBLFba6S88T0dJ-vLy6SMm3OLIMEdllDfVI,6663
245
+ sglang/srt/layers/quantization/base_config.py,sha256=jWk_egQrVNMYmQgbTI9vkcgzScLFjB5_sywFlAfE5J0,4776
246
+ sglang/srt/layers/quantization/blockwise_int8.py,sha256=BS4nmo18QaC3vyCe9Wwe0Te-7FmFUe-udl7xbhRaU0s,14887
247
+ sglang/srt/layers/quantization/fp8.py,sha256=J9P6SwZ1PIb3dEJoP7X2Uw5VOtU35vU5jugkoYyaZC4,41081
248
+ sglang/srt/layers/quantization/fp8_kernel.py,sha256=JRalHJ-btDpzl3oXu2R_ZoJBu5TzBBmW_wKZDFs-usQ,24384
249
+ sglang/srt/layers/quantization/fp8_utils.py,sha256=OL_tV-NvNGcXV2i3mOXytjzi9d4t59vOxvZN_hBc1KA,21572
250
+ sglang/srt/layers/quantization/gptq.py,sha256=e4rMz374-yQQqeAI77WPxfcAaRk38GeN2akEpvnC_Do,15141
227
251
  sglang/srt/layers/quantization/int8_kernel.py,sha256=GfRn_imIw8kNgqdtb2lr7BettjgDgimbl1Rubnamjh8,11352
228
252
  sglang/srt/layers/quantization/int8_utils.py,sha256=YK9CS-lb_n91kNCTKK5o5apYF31V2giDg5G5VKrpcUA,2356
229
- sglang/srt/layers/quantization/modelopt_quant.py,sha256=F16S-ky8EiqBrTU-u3gCmyc__TVPu03bR1Xo8ojCzRk,6989
230
- sglang/srt/layers/quantization/w8a8_fp8.py,sha256=AOczkwzh6gRx6nvYFqbrdNyAxmlPsXhws1DEX_nP0Ss,3672
253
+ sglang/srt/layers/quantization/kv_cache.py,sha256=rJi6amyLZsquUMo_V5iLlPMqdsGTLgxh4popN1xUHCQ,4236
254
+ sglang/srt/layers/quantization/modelopt_quant.py,sha256=mne4uKF0R-K0OvWN7X5ZxD4LdXKBc6GvmpZzIW6gkmM,6969
255
+ sglang/srt/layers/quantization/utils.py,sha256=BmbovB_SxPcV7DRzGod8plawLbWRLIiN8Q2l_oWtAfw,5627
256
+ sglang/srt/layers/quantization/w8a8_fp8.py,sha256=XcQdgqXA3eKbAf-4_0I81Y5Nvjns3bQTocovnN8141w,6234
231
257
  sglang/srt/layers/quantization/w8a8_int8.py,sha256=V5vxn0wmUL1szj38lsJOKeNNEvFHisU7hZZLO4FfoNc,8733
258
+ sglang/srt/layers/quantization/compressed_tensors/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
259
+ sglang/srt/layers/quantization/compressed_tensors/compressed_tensors.py,sha256=YEtWykakVRjS4rko8T7ui0uf_Q15n9SPcASY9EkWYGk,25344
260
+ sglang/srt/layers/quantization/compressed_tensors/compressed_tensors_moe.py,sha256=jHuFgm6s0xxijtceDySj-myyWVn2JJMx4ZXn-4P16F4,24985
261
+ sglang/srt/layers/quantization/compressed_tensors/utils.py,sha256=mnUmKWFQUnY8bVoFHUuNVwqsfS-cefeR-ofyaihCXcY,7621
262
+ sglang/srt/layers/quantization/compressed_tensors/schemes/__init__.py,sha256=qcdRgoUNJWXqSimns-D987TW0OTk3uFuWNBX8Z6H8Fk,246
263
+ sglang/srt/layers/quantization/compressed_tensors/schemes/compressed_tensors_scheme.py,sha256=tdKJC8c3SX8T3z8JL-1YCsg4ftcv55Wxt0vZrYftpX8,1635
264
+ sglang/srt/layers/quantization/compressed_tensors/schemes/compressed_tensors_w8a8_fp8.py,sha256=BSgSFFeuBHSnZxvVpNDojeGCx0ClDDH04aTjdBUyE6U,5940
232
265
  "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
233
266
  "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
234
267
  "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
@@ -381,127 +414,139 @@ sglang/srt/layers/quantization/w8a8_int8.py,sha256=V5vxn0wmUL1szj38lsJOKeNNEvFHi
381
414
  "sglang/srt/layers/quantization/configs/N=7168,K=256,device_name=NVIDIA_B200,dtype=fp8_w8a8,block_shape=[128, 128].json",sha256=PD4AJYCkHfy2ivv9baMouFXzBTy0eKMumbAfxfm91HI,3256
382
415
  "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
383
416
  "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
384
- sglang/srt/lora/layers.py,sha256=r34oprzwyE3SWPvaNkBvXWPtfa-0IY987_bjj36ySfw,9996
385
- sglang/srt/lora/lora.py,sha256=07-IaAfbb3zPJ-DukHL3uyQ8fjJx_hrVFHS2pqe8LZg,8238
417
+ sglang/srt/lora/layers.py,sha256=eqU5PxLx9jsmp0fxQ-e9mlSD0Zz4Y9Uan_x9Z5-y1EQ,11835
418
+ sglang/srt/lora/lora.py,sha256=9BpasJObx-XL_qSoQhheGaYBJ3PlKqJAARvruIOJE0w,7514
386
419
  sglang/srt/lora/lora_config.py,sha256=qDgMTx_69jyJUl29O5FxLzYa0BMhqYVXWXfyyVOvGm0,1684
387
- sglang/srt/lora/lora_manager.py,sha256=Wlq5dxLM7Uj4uTGpFXH1q-IOI8j4mFXYHPKSltx2QMI,7794
388
- sglang/srt/lora/mem_pool.py,sha256=eV_GXETxNODPVIAnTEeUUUVn0IVgguBR_mYFzIK-VHA,6835
389
- sglang/srt/lora/utils.py,sha256=6i7Q1Y-1LLbRkeCMv_lKIzkTN0veUTLbc8wlHn7R-bA,4571
420
+ sglang/srt/lora/lora_manager.py,sha256=Gpkq4N_cJGMIDtxUCScwP4LGcHyUJZ457EI_ti30_A8,9187
421
+ sglang/srt/lora/mem_pool.py,sha256=xUFoHUDJgX9lt2YugD9HUY5tIMnJiazYMZ6LYqSGv-E,9633
422
+ sglang/srt/lora/utils.py,sha256=GjEBgsGhDhX4NqVqeaciznQ8RotKZmb2c-nw4YMLHxA,5251
390
423
  sglang/srt/lora/backend/__init__.py,sha256=FziFT8HguMFj-h0tUCc4_UEbtOWMlYi4gNlYJcArWh4,671
391
- sglang/srt/lora/backend/base_backend.py,sha256=dldwA7vTWrB1ln1MwLYKNtMkBoAgD7OLSlWe9tL2lzk,4602
392
- sglang/srt/lora/backend/flashinfer_backend.py,sha256=fXfkl7Cpw8ap2bCrgWdn_gEUzMXX1pNjNuiPw3kA76U,3984
393
- sglang/srt/lora/backend/triton_backend.py,sha256=ZT5M30vj8x77Kltukpga4wk1sd8fT4n_FdsOMQBTMI0,2610
424
+ sglang/srt/lora/backend/base_backend.py,sha256=tGpABAn3DVC8GONf8USkaxkzkpVsDYfgKrnLCsXpivo,4558
425
+ sglang/srt/lora/backend/flashinfer_backend.py,sha256=VmDSY2YqTLK2EBeqdMiNoirPxDifCMmfiCB3HNwpgvE,4138
426
+ sglang/srt/lora/backend/triton_backend.py,sha256=_QbqggFPHMPWgx3PI15yyyfxBCTxSpGA209x_frI12E,2517
394
427
  sglang/srt/lora/triton_ops/__init__.py,sha256=JGOYPIn1XbGcyJTbt8A0qoc02PYONSGNNjGkC8yJpAM,283
395
- sglang/srt/lora/triton_ops/gate_up_lora_b.py,sha256=qve4oNZHYUFk9ckmT2BVuDNMEvrN7Quu6RsS8Iz3uRQ,5066
396
- sglang/srt/lora/triton_ops/qkv_lora_b.py,sha256=BmIcTZMnlSnie9rnMl4KvLpc4Njsk7_IppbUqitf9Xw,5738
397
- sglang/srt/lora/triton_ops/sgemm_lora_a.py,sha256=kv-AvJ_Bi3yWjGvFnSwXvP66iJvY9n9pEnJzJ9-DWzo,3982
398
- sglang/srt/lora/triton_ops/sgemm_lora_b.py,sha256=Ai5vPriT4OgACwK7xrpGgf5L1oaN9x0jwNKMChu3uI0,4299
399
- sglang/srt/managers/cache_controller.py,sha256=XB6UHFXVw5LAyKV2jK0iVzhAYT4fSzJVCNL42eKGUsk,18175
428
+ sglang/srt/lora/triton_ops/gate_up_lora_b.py,sha256=CDGt7lpu9GjykgMtmwbZ3PEqjTlRYyh28AUlj1cRcmw,5279
429
+ sglang/srt/lora/triton_ops/qkv_lora_b.py,sha256=HTfU3HxxxVyaG_aJrrVjPJTnqf62yvepcKJKYkG0XJQ,5944
430
+ sglang/srt/lora/triton_ops/sgemm_lora_a.py,sha256=ZmWEqHJaorRNNj-c_ZXPi_pX8X_yIAwudRHAJVa0m08,4350
431
+ sglang/srt/lora/triton_ops/sgemm_lora_b.py,sha256=Q58UzWUb3QFqY_ZxWA3poN373N0Hwkks5AQRKIuvFC8,4517
432
+ sglang/srt/managers/cache_controller.py,sha256=BRRyu110lJalLVLZbTCMOWbD4s-LG3f9Xz6sxU2dCrc,18144
400
433
  sglang/srt/managers/configure_logging.py,sha256=fOJaXAQ1n9m-8KPJndpsKvS885i69SMafoEADLIVfIM,1633
401
- sglang/srt/managers/data_parallel_controller.py,sha256=px1SrFUUDDroXaHN4m9LcMNqGDt2vv9COiVNg1O-BqI,9561
434
+ sglang/srt/managers/data_parallel_controller.py,sha256=Xkj2n9uDyq7a-AVDZlfzeuNkC4ibsSftb1_bed9hgQ4,10318
402
435
  sglang/srt/managers/detokenizer_manager.py,sha256=HTfpJWMF1EImhKOnLJ96xPmYXm71xzaisLMfxg3zpgs,10111
403
- sglang/srt/managers/image_processor.py,sha256=o2ChZW3mWSw1I_4HFE8SV1ZsYtWp9G9w7Ij73BvugN4,1711
404
- sglang/srt/managers/io_struct.py,sha256=XvXM6WRarwEtaNQWlzZpYKy2Zi4wCG3xnz_ojskYJjM,22791
405
- sglang/srt/managers/multi_modality_padding.py,sha256=dkFKknxML8R3KrAZhKXO-A69u3pUUI1pRfOVMJfqHYw,5165
406
- sglang/srt/managers/schedule_batch.py,sha256=NzT2ivSSKRqtFRiAr-d2MO6spH0z-cf0jANw3JyGrr8,55983
436
+ sglang/srt/managers/expert_distribution.py,sha256=r3o5RGI0gnV7xb60AApqKYa0oiSB37oB7hQBX7P3xZM,3225
437
+ sglang/srt/managers/io_struct.py,sha256=_WYQ2x49Wc8WqmZH0Q5Te7zVrGRQkbn0ADghuDwyk7k,23852
438
+ sglang/srt/managers/mm_utils.py,sha256=KzodrStj3ouIEHLRja3TUml0YUQ59qmEPejks5ikPQk,13828
439
+ sglang/srt/managers/multimodal_processor.py,sha256=nTlktTXYo2NKr5Ab2AaSjydNQDdwpJHs4XdYr5zdtkU,2154
440
+ sglang/srt/managers/schedule_batch.py,sha256=2HOloc18BAM10aL8os34sQ3KM3sYi-yhDDahlTxrBtU,60106
407
441
  sglang/srt/managers/schedule_policy.py,sha256=E1qVq2G3jptKdX9nlqfayeRBUll9xB6bK8nBf3EW32E,19469
408
- sglang/srt/managers/scheduler.py,sha256=bL1HiKey5ZgyqcYFur6pfEZLbBPcKLF-peAreiNnfK8,69433
442
+ sglang/srt/managers/scheduler.py,sha256=kH8oC7R_gir2Whsq2NMCbzdmd8nqGx0GXqXuVwr8dTE,80004
409
443
  sglang/srt/managers/scheduler_output_processor_mixin.py,sha256=u2sj6MViFTov0lVZSysZ-wph2pEqRCtCjwA1UdttZ7I,26338
410
- sglang/srt/managers/session_controller.py,sha256=YOv8cFcuVmdCE4OfQJ6aA5AosHRwtZF9WdUUfUXEp0I,5753
411
- sglang/srt/managers/tokenizer_manager.py,sha256=_xIn-3BDgagqnmRKFhIQTkCg2jXk6jp607B7CeprNVw,44686
412
- sglang/srt/managers/tp_worker.py,sha256=o9MY1a8x81nI3W0m64YvOXcAA4sCBvPPmnNU67vBXGs,8710
413
- sglang/srt/managers/tp_worker_overlap_thread.py,sha256=BhbVjI4DDXBIbRs1xzxME5uHy6WAxmZSJBepRu0Hv-0,9108
414
- sglang/srt/managers/utils.py,sha256=5i75uLlQOF_5CaT02CrWtwozMTtwTg2_nLP8Dtr-JZQ,1536
415
- sglang/srt/managers/image_processors/base_image_processor.py,sha256=6UdYo3lFgrMsn-5a8NDKhkjeJ9bxPDFRVzhXv_wV5fg,7567
416
- sglang/srt/managers/image_processors/janus_pro.py,sha256=2Qmt2gKPT4-pUnTBhEEokiBEQ0kEjdZjl0Eq1U6y42U,2615
417
- sglang/srt/managers/image_processors/llava.py,sha256=UdNfmAZuw__MgySC4uu5NsWgBXPyP4V4lIzDEFV9AHs,5839
418
- sglang/srt/managers/image_processors/minicpmv.py,sha256=JS11rO_gyRVLijD2AxNM9ZJZOFxLbver_v3bqqUEhBg,2881
419
- sglang/srt/managers/image_processors/mlama.py,sha256=ECN2rmS2Vl9eg3u-XmwtphQm7zTaB4clK2HU5ntuODE,2186
420
- sglang/srt/managers/image_processors/qwen_vl.py,sha256=wJbbCW1DKPQp_RufQ4fz7Nd_hRldfwe97e_2YMUGmJI,6417
444
+ sglang/srt/managers/session_controller.py,sha256=o-ifit0n4_xHLNmyD0Ams8FxGRgxFybX-Vz1hwgr3UQ,5755
445
+ sglang/srt/managers/tokenizer_manager.py,sha256=cYzNwHKr5Y5LMwHeU1YIyTIR88-QMpMYqcWt1nm8KLU,46161
446
+ sglang/srt/managers/tp_worker.py,sha256=IFiOhbNIya-7cqgp_Yg0ZXGcsgy9YS295AfxJYjFqzQ,8833
447
+ sglang/srt/managers/tp_worker_overlap_thread.py,sha256=3_ZJ8Rq7v2ZDaRNTRu5Dy8AbqiAlJQp3IAKnn_WAwd8,9127
448
+ sglang/srt/managers/utils.py,sha256=59IuYvuEfi8ji_acKqH3Y1-2PaA_dSlZMbHSCqjQfCA,1629
449
+ sglang/srt/managers/multimodal_processors/base_processor.py,sha256=PpNwQBhEuXbztzS3cBFu7UL5sfKYay-WoqAweSVRRao,9984
450
+ sglang/srt/managers/multimodal_processors/clip.py,sha256=cuC7jynTkqRAm_yixey0Tv1KSblI97lCzOuX8-Ix8sg,2148
451
+ sglang/srt/managers/multimodal_processors/deepseek_vl_v2.py,sha256=-qU99yCs3TYM8g9vOo5_SIWX91theu0ZREXZT5-A9yY,4594
452
+ sglang/srt/managers/multimodal_processors/gemma3.py,sha256=Ic8wTOS5TS9qc-SsGkKWQ_7NgnimCnnBeXw4UB8DTgM,2745
453
+ sglang/srt/managers/multimodal_processors/janus_pro.py,sha256=GaVH3_HxGcQHRvFuP5zpkQ9m678cyVrbjb1iwwdV_fE,2780
454
+ sglang/srt/managers/multimodal_processors/llava.py,sha256=2UfMVl-CoQ00Wt8XCZLkE3G7mFdU4Ol1aid6PPG-ZXw,5711
455
+ sglang/srt/managers/multimodal_processors/minicpm.py,sha256=eD-FgRMTSF9jvolMQYfIgzt31qs0SyLEcVrpHaf46fU,5979
456
+ sglang/srt/managers/multimodal_processors/mlama.py,sha256=K-_guz4aaxR8R2zm9MAnLy7K66fkd_0J6R9P_4wdIb4,2118
457
+ sglang/srt/managers/multimodal_processors/qwen_vl.py,sha256=OBSKa6LGhTl6wJJ24Bk0IibcKM97Q-6XeBJ1jxLGMbY,6654
421
458
  sglang/srt/mem_cache/base_prefix_cache.py,sha256=NY62Zo0A0tLJ7ObRLOQqQcXCxoJUDZsK8f5U4dNQjKc,973
422
459
  sglang/srt/mem_cache/chunk_cache.py,sha256=it5SfL1FwMbrdeOH-I-Eu_i-I9hFB1xL-z_brIUoCkk,1835
423
460
  sglang/srt/mem_cache/flush_cache.py,sha256=GYcxmNXh4hsMpFfNOuCTpKilW7guZwTtAg_usVeM3J0,979
424
- sglang/srt/mem_cache/hiradix_cache.py,sha256=UTdIqGosuU3PjGj27hmLuf2exPd9Rkg8GShVqi85FUA,15548
425
- sglang/srt/mem_cache/memory_pool.py,sha256=xZc9pXgCwQyfv7rkgh4uoXDi2qPwueqTuw48poSV4OQ,22816
426
- sglang/srt/mem_cache/paged_allocator.py,sha256=DRq7qZXLVeHKkCWRlEkoAmLP5IC_1ReHa_xL7-8Orzs,8594
427
- sglang/srt/mem_cache/radix_cache.py,sha256=bLldgkFZvmFOVgc2xLI87an3UBYxIsOlRBsxenFEL7s,14912
461
+ sglang/srt/mem_cache/hiradix_cache.py,sha256=BjPZXHYmCW6Sdohb67Qqzf2T0YXkO0hEPyynvn65xHs,15956
462
+ sglang/srt/mem_cache/memory_pool.py,sha256=xfEFCYs3DsW7t5Nuv9pwFKtap3SJgxkvXToKQFUT-rI,28255
463
+ sglang/srt/mem_cache/paged_allocator.py,sha256=Fl6d8rgkwGIgq3n7AKM7Jmm_aIgu86jJ24636nIkX6s,8594
464
+ sglang/srt/mem_cache/radix_cache.py,sha256=Lm-pco6CJ4orb9IfDpbHm5MnyK8Ya0OF1x9p88dv548,14906
428
465
  sglang/srt/metrics/collector.py,sha256=0X40ZZ18182sx2t0eqeqoK7gspH36L343zNvSkgBvd0,9293
429
466
  sglang/srt/metrics/func_timer.py,sha256=VFyNRrbnKVCwnQsrlLin1lITJfjQpf9m8sGPqL5LIsQ,3438
430
- sglang/srt/model_executor/cuda_graph_runner.py,sha256=EllAbxzJy6ULgJyboVYmqG1IFYhNM6zLAAmpWARsIRo,22197
431
- sglang/srt/model_executor/forward_batch_info.py,sha256=wtByWmk7u4YUH2ZergdjT1tyl5Jhc75S1cVvd_GTEQI,17404
432
- sglang/srt/model_executor/model_runner.py,sha256=zE3nWMBLI0eW2mDZHBvxj1bHW3VsAxtd-RP6uddyP5A,41581
467
+ sglang/srt/model_executor/cuda_graph_runner.py,sha256=rIxQ4Yvt5-i5oDBxkzEm3OtvWid7yx73jUtMILF6WWQ,23207
468
+ sglang/srt/model_executor/forward_batch_info.py,sha256=RKRQMoMcAexG-YweQKl-uSXc4qe-yk4u6iTtWK_BOpw,19591
469
+ sglang/srt/model_executor/model_runner.py,sha256=TMcejBnQxNSZGst-8B04z9IX6smVxCjTeA6dy2-5U84,44807
433
470
  sglang/srt/model_loader/__init__.py,sha256=zGZkOBz1zx-pkaIy47BasL3fjDlAcxAXUTjInOhXHAE,919
434
- sglang/srt/model_loader/loader.py,sha256=bHsb3T9CGac32NgDeRt03cJXBy-t1WsyYOWcYkeMVLA,46813
471
+ sglang/srt/model_loader/loader.py,sha256=mK9ZspmhUhw5A0e5G3DF77HJ4-JsDvzLvwj5I2J3Vdg,53883
435
472
  sglang/srt/model_loader/utils.py,sha256=0NaMR67fESFopaklmsleiL27XH1QUrjZW246MUu1EJ0,1369
436
- sglang/srt/model_loader/weight_utils.py,sha256=26fCuVipwEiDjziwsuavwa_BC1zPB2-socfly6IjmG0,30373
473
+ sglang/srt/model_loader/weight_utils.py,sha256=Bkfgz6TUjkJJb8BiFxrv7FgbZFd9eW2y21jDBPdLWSo,32056
437
474
  sglang/srt/models/baichuan.py,sha256=iXgta-W38OWesxmXWZJ73fUvPdu51EwTQzUD5mmfJ8s,15721
438
475
  sglang/srt/models/chatglm.py,sha256=avLC7mjjGskBxCxy-9s0sMlAJjfFoG_y8VieR1QfDsM,13918
476
+ sglang/srt/models/clip.py,sha256=vNKS-wWs6Pl8hwJUY-3g8L3uax9Z_Voc2CF0C8IEEiQ,19554
439
477
  sglang/srt/models/commandr.py,sha256=Ug-B0QcdWZufrTybC6K5yP3MLKNsYb-vzfrqUsXYUcI,15276
440
478
  sglang/srt/models/dbrx.py,sha256=0Vf4yhqe8YeQuKR3P-agvYJScmHwH3-tFbyU8kv5QJM,15559
441
479
  sglang/srt/models/deepseek.py,sha256=Le2MXij8m4hT7QYgD0bFMFmYhbgX7SMjoXZFB8BxgyA,16871
442
- sglang/srt/models/deepseek_janus_pro.py,sha256=sduOYeAQrb2DFfoeJVzxMNOAgS7_CmTqaKWc8J3Ypeg,72528
443
- sglang/srt/models/deepseek_nextn.py,sha256=8uR4k1qCBKL1JAcE8cCJz3HowrIgCgdMtbqQ4Xq8Tr4,13161
444
- sglang/srt/models/deepseek_v2.py,sha256=oEREP6cErJxBuQRooXY6tZ7oMwLvWLTjDYZPAfXVAS8,48651
480
+ sglang/srt/models/deepseek_janus_pro.py,sha256=D5nJgMmrIqXW71r9_-_yd63G8DftnM1K8mS4hU8O_pc,70547
481
+ sglang/srt/models/deepseek_nextn.py,sha256=kca-2Fm2_SmqbOEFfd80pobooi1BXd1oe_4EsUM6SeI,13561
482
+ sglang/srt/models/deepseek_v2.py,sha256=7ljI2R9GmngO4gcrV3MLmioxeEKgCFyG-YL_J6QYA8Q,58695
483
+ sglang/srt/models/deepseek_vl2.py,sha256=T7zdLBnnBSVPCX4cec1B-MbnXru1-70KqEdDLqbYb_g,12956
445
484
  sglang/srt/models/exaone.py,sha256=5iibqQTjpgosuGRt2rj2lWR0ShK2XGhbdFSnOWpaQss,13386
446
485
  sglang/srt/models/gemma.py,sha256=3XxMDOKz4xMP6VzWoW8f0hmMf8LP8fhzMw5prsYC4e8,12602
447
486
  sglang/srt/models/gemma2.py,sha256=MDe_HNkSpEJpw426tbx3fp271GBlSVEuhIdGeOB_jYA,16356
448
487
  sglang/srt/models/gemma2_reward.py,sha256=V8U3_ADUHWPdOwvEe1jhGW-oJmBgL8t1TY3-67Ksv2A,2618
488
+ sglang/srt/models/gemma3_causal.py,sha256=nKO-DRtvcXn5bHquxILgnp0fJT6usoB81W8kPVdYsA0,24934
489
+ sglang/srt/models/gemma3_mm.py,sha256=ZFgHB9-GBNZcMtYvUdynk16mWmKQ-oYxkDHhttWNQB8,17854
449
490
  sglang/srt/models/gpt2.py,sha256=dAnfmsAL7JVHakryqrERR1jgL8mI1Op6nPHYfDCF7Ao,9802
450
491
  sglang/srt/models/gpt_bigcode.py,sha256=EAN6xAXpa8m3DcBuH1D4rTPji2oG9NSozGXSNHtE2lw,10268
451
492
  sglang/srt/models/granite.py,sha256=nu_Zl_PYn188gk1uYVZ76y4wwHZV7G0w7uanhqpSFUs,20813
452
493
  sglang/srt/models/grok.py,sha256=pQOXtpHOYVntwt5QQRLffYsnMHmMfPMmGyKMfR0k0Ic,27994
453
494
  sglang/srt/models/internlm2.py,sha256=4eh9WVgK4yg13IsnH5qB2xUCWnixj_aLLz7qa_4m2_Q,13017
454
495
  sglang/srt/models/internlm2_reward.py,sha256=ndfGmyqYZbVZ7C7rJ-v9oK3wa-EpoBGybS8MlyKZi2E,2522
455
- sglang/srt/models/llama.py,sha256=FEtMrzjrbQLOOQvrqGFwslizjHPaU--9rEiXlr7MIiI,22958
496
+ sglang/srt/models/llama.py,sha256=0jD_xcCuLfDGfHrzwn3oRZvRxilaWb4ARasu5U1x9zQ,24529
456
497
  sglang/srt/models/llama_classification.py,sha256=4QWTFaUZIFKYZvEzs8bx8VkOZNIwdYCLrnwrdAw4QK0,3108
457
- sglang/srt/models/llama_eagle.py,sha256=4ynCbF2Lp-t0_T036N_Pa_8a1SIPkNeWuEwF7r5x1ls,4819
498
+ sglang/srt/models/llama_eagle.py,sha256=OB2lKsjn7BcfCZljklnhk83me8j0PuQmYLou7baNcq4,4866
499
+ sglang/srt/models/llama_eagle3.py,sha256=v3bftBVDIGjnzngQYnu19cy0J_3w7yruHqLP5nsAQDM,6642
458
500
  sglang/srt/models/llama_embedding.py,sha256=zq-_lNu35VBFc7eemiam0zdkGIE8fzrgk5OWYfirZnA,3254
459
501
  sglang/srt/models/llama_reward.py,sha256=LF2nqMV5XOrljGjAwJg43mBv3z6Q040I2EYlgZeCp8k,4681
460
- sglang/srt/models/llava.py,sha256=J06XzASrhg2Pw1Z_jMoC1lXI4hFfoZacqS2mhiLI41k,26778
461
- sglang/srt/models/llavavid.py,sha256=iwqwTruJTG9D1zV95RHq-RpTp1g12kKPrNIfv2a2XyI,12485
502
+ sglang/srt/models/llava.py,sha256=G3aaWr0Pldx-yqmbv33KMMwpJ7TAW0yIY-eE9bFfWto,26785
503
+ sglang/srt/models/llavavid.py,sha256=w_eaZRSObvqftKE_s0cAE5NvUyFtHfU1EiD0XFBfSow,12492
462
504
  sglang/srt/models/minicpm.py,sha256=-ot45U_Bv4x85JdbIAQXoxa1sF-ZDkBk8flU-Ruli5Y,14652
463
505
  sglang/srt/models/minicpm3.py,sha256=sRHPFUH636GIY94B-hpAN2MSzYT1pzLPVypTNjUtttY,26270
464
- sglang/srt/models/minicpmv.py,sha256=AYxGTFbvfhIgmGmM7ab6620jtCS2qabh5tJ6NrkQ_0U,43839
506
+ sglang/srt/models/minicpmo.py,sha256=b2TE13gXgfcBg9ibYaIHbFFtsMbPahs_qJC01BaMKu0,79521
507
+ sglang/srt/models/minicpmv.py,sha256=cMKLdprNl9siOlHJDY9OPRRDBCax84zHWCj16m-kzaA,40658
465
508
  sglang/srt/models/mistral.py,sha256=EYifJUUzN2Z2-iL37eJiNZF_DB0H4pa0mKlgYRIxM70,838
466
509
  sglang/srt/models/mixtral.py,sha256=6Fse2J-20IMylP-yzpEihIinaH37TmmslATbLcWBRYY,14926
467
510
  sglang/srt/models/mixtral_quant.py,sha256=MSa6UKPbgv8Rn8Iv8o1dQhcstAHLNQzE0eepFx_hYSw,15221
468
- sglang/srt/models/mllama.py,sha256=jG29hG_O6YzzakDP48hITQdMiNKX6fapkQ0duiyAD90,37887
511
+ sglang/srt/models/mllama.py,sha256=1hJiW6AHakH3QvGC2zKxuvZfo6YO5fbAmp5W-i2swB8,37891
469
512
  sglang/srt/models/olmo.py,sha256=FJk8A3T3TF5QcTV6rMP8np94QtvxpMWlgCsv_5VwpVE,12632
470
513
  sglang/srt/models/olmo2.py,sha256=U0ScFzWazOrb_Q90sfXkpVNAsXT-pgZbNgGh80R40VE,14288
471
514
  sglang/srt/models/olmoe.py,sha256=tx5OKWLOr6_pohe2eBcIodCmcuSjtpteHq_tG_QVYCY,15910
472
- sglang/srt/models/phi3_small.py,sha256=_ZDXVJN3B5f-46MMKxNim9874cVqJpzBipbmfBz6Hn4,15464
515
+ sglang/srt/models/phi3_small.py,sha256=6p-5EBbwN3FmhoL0VNrNb6VP7HqIhYgVBfQ98L6Rjjo,15469
473
516
  sglang/srt/models/qwen.py,sha256=edS0UYq6AoHZdYUJtQa5wyFNzZMW0JAMmBulH2uheaw,10719
474
- sglang/srt/models/qwen2.py,sha256=_jo9Ox7MrXCicw7Lhjla29UtRSXfsNd079So81rNcLM,15947
475
- sglang/srt/models/qwen2_5_vl.py,sha256=oA6Cz3-3RQN-D9B88s8dybV8idGjpdu2Fcm60i77Uqs,28184
517
+ sglang/srt/models/qwen2.py,sha256=D7VIFWbQ_MxY42K9j7ksgNTP9TNh_FA8u4FUND2xSuk,16038
518
+ sglang/srt/models/qwen2_5_vl.py,sha256=iZK_giWA58sb7cK9u_T-F3YgdgwDBq79nl2CszTL7Uo,24506
519
+ sglang/srt/models/qwen2_classification.py,sha256=dGrMm4ebd30_lBhHOhaV57ig2iOTx3nqB4GEzsrRIM8,2747
476
520
  sglang/srt/models/qwen2_eagle.py,sha256=Iz0HWL2FgSD3FqoFhfYmbIZeEYkPTJ96lYbkncmHJX4,4644
477
- sglang/srt/models/qwen2_moe.py,sha256=zYLJecN1mUyMBmnZoVaJd8LUKT4YZPBIO1lfFOqmU-c,17755
521
+ sglang/srt/models/qwen2_moe.py,sha256=9cLOPHBpwdID92Ed1CEjMUxMxcWX83lklunHbyK4To0,18206
478
522
  sglang/srt/models/qwen2_rm.py,sha256=-mQXDEv11p-I1HXgYLTtY6ROem6UYorO958WsDrzsgs,2837
479
- sglang/srt/models/qwen2_vl.py,sha256=G1Hx3n-5pXRtVHEtS-UjiQTr4AvltXwQ4PyroxNGfcE,23918
523
+ sglang/srt/models/qwen2_vl.py,sha256=JlcMUKKi2ynZR3vQN5gAwL13KEowq4Irc3Vj2WGpdnE,21983
480
524
  sglang/srt/models/registry.py,sha256=inKh9iwOp3LFYm3nqujg-OtABClOP-ifc1stA9cZegA,3434
481
525
  sglang/srt/models/stablelm.py,sha256=w93fNXpDwQbuKi4tdeo0bsXFZrMZVY4_pgNL0E5RErQ,12242
482
526
  sglang/srt/models/torch_native_llama.py,sha256=5tfFSMAXB3ScToqTALtCXa8Oo-qPCJh-KQCNB6QOlNA,19293
483
527
  sglang/srt/models/xverse.py,sha256=I7ivNsk6NRqPxlMUmdclpzDCvhAnWbv_GOj01MKHJrQ,13996
484
528
  sglang/srt/models/xverse_moe.py,sha256=xLwn5pRwQrvj7zMmwl3o49m7xILb2ACRdWvm9hY8LDc,16743
485
529
  sglang/srt/models/yivl.py,sha256=oToK7-u5IGO7xwpJIQ7VtudlK6-zPqJX4bt6_wv0SH8,4850
486
- sglang/srt/openai_api/adapter.py,sha256=GAAo_Vuf9aiKgUH0NPanDTB_eC1WqRRNnltVvAFeW4s,67412
487
- sglang/srt/openai_api/protocol.py,sha256=6-vwYewe7JfmMMC56oRU13aRmmKUerZs50GPmiGeX7w,12490
530
+ sglang/srt/openai_api/adapter.py,sha256=q0364qVBd-iymVnVBngv4ZMdZorl5XEYN6u8ZZCaDTs,71454
531
+ sglang/srt/openai_api/protocol.py,sha256=Y8PFFhLbzhpoERM6-WsTkm-ZuGcE-3tfenh9e-AC1vc,13374
488
532
  sglang/srt/sampling/custom_logit_processor.py,sha256=tDvoLgLqn-sy1qcY6vSrpbnHCeqbdk0uhMOO-uy4p4E,1099
489
- sglang/srt/sampling/sampling_batch_info.py,sha256=T4UJ_CRB6A0HnRwEsqmxy1CJQMwZaVdxbdlCeTK4BUI,11992
490
- sglang/srt/sampling/sampling_params.py,sha256=HihGfhdR4FmOeltEqpW2kSLfNu94VCd8l0RNOQFSl-Q,5919
533
+ sglang/srt/sampling/sampling_batch_info.py,sha256=wrGGU27mWOi_yCBBCOvTQLBdyTjfkPuj7Hsk0zOFyH8,11989
534
+ sglang/srt/sampling/sampling_params.py,sha256=nXm44Inn91YtrMpAm5mDb6-97owRy-Bh6lZ0BIpw73I,5919
491
535
  sglang/srt/sampling/penaltylib/__init__.py,sha256=mtN8grFEcaBUhl4yBHmw8NNirt_i6uKO2cDNLHOpZQE,496
492
536
  sglang/srt/sampling/penaltylib/frequency_penalty.py,sha256=Loc3qjJTksNc5s-DV7QZHjgqoo5pxk7-nZzxwyhD2tQ,2144
493
537
  sglang/srt/sampling/penaltylib/min_new_tokens.py,sha256=rdU_D7RoIcrQPhysNQEzmr4TO2OoEi___p-i3QdwkgU,3331
494
538
  sglang/srt/sampling/penaltylib/orchestrator.py,sha256=XM-Lm6u7gYPtMZrTIc0FR4QxNZxBH5s_Cj82umyCzYk,5721
495
539
  sglang/srt/sampling/penaltylib/presence_penalty.py,sha256=NRh10AJrrQlGJ6S-enGdRefrTrWpyqrSm-aNnyqQNQQ,2119
496
- sglang/srt/speculative/build_eagle_tree.py,sha256=8_uUpkQAE0qcn5mA6NPjfl81EMuNxg4fZq628wjEZNU,20805
497
- sglang/srt/speculative/eagle_draft_cuda_graph_runner.py,sha256=zsF4AcsykaFjzZ5SKdvUQyEB1GVXTvR1_kI_oaQafBo,7704
498
- sglang/srt/speculative/eagle_utils.py,sha256=DhpNNNN56jw0ucXSBYq9_IzDuJmEKgXS6lbLczaghNs,27196
499
- sglang/srt/speculative/eagle_worker.py,sha256=ewqd3NxP0q8CHIhsigHYATv1SUnVl6SU7sUo5aLnFhE,20437
500
- sglang/srt/speculative/spec_info.py,sha256=hJR0b3gZ0QA0KZLq6TfqSDJkmpV6mThyle3sHYI4o4M,522
540
+ sglang/srt/speculative/build_eagle_tree.py,sha256=SFQ3eHbhfNxOdxgqDP5wSV_ZlIVqLw7VivycNZ963N0,11690
541
+ sglang/srt/speculative/eagle_draft_cuda_graph_runner.py,sha256=t2mbGZY23E7WraJJQW_4uaLTfnbPDExRVC7jldtuY6s,9287
542
+ sglang/srt/speculative/eagle_utils.py,sha256=_T3eRd59bzCR2YE_Z4W6Pux40KpcueWwN-rrLe-ikFQ,25812
543
+ sglang/srt/speculative/eagle_worker.py,sha256=W6hoEW0tpNaoC0T01vEEMJOwaiZjhH4rtbvd2FOKxiY,24229
544
+ sglang/srt/speculative/spec_info.py,sha256=rhaKG0TzyF9XZYHEWp1jccwTBohSNsUDvxHFtAoOl18,709
545
+ sglang/test/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
501
546
  sglang/test/few_shot_gsm8k.py,sha256=7VLbWl4nCQs1wjtW4q-46jf9jUCycSs5Iw8v7sUSzBw,4284
502
547
  sglang/test/few_shot_gsm8k_engine.py,sha256=QQbrwOX6-cJDD3RZC_e7zPnt6aSo8JdF8X_lRHSjdDM,3886
503
548
  sglang/test/run_eval.py,sha256=9yO0hXZOcn4abEOs96T-XPguDEklK16Ltco0pGF3zCg,4020
504
- sglang/test/runners.py,sha256=4ZH0AtzMNFzxLXCxJlKpQOLl5C3jkut9YkDAvGYxW5w,29191
549
+ sglang/test/runners.py,sha256=zl_7wdwUbuCSZoA-f94VhnOI36VX_DwCt3cAEzIjm9s,30484
505
550
  sglang/test/send_one.py,sha256=6FhbJ3c8RpXxvFTELRXaF97GpT7zXXsCDYZh1DqG22E,2550
506
551
  sglang/test/simple_eval_common.py,sha256=joqrGysuLnJFtzDRIgFkMsRyKUSyjVPFWp0_PHAL3Ik,12378
507
552
  sglang/test/simple_eval_gpqa.py,sha256=8Xt9Bw05c7SZTYrCZgB68OZUqUbLo69ywiyx0bTvSUk,3220
@@ -509,15 +554,18 @@ sglang/test/simple_eval_humaneval.py,sha256=zmV3xWYc2OrpiT9Dy55RTKZL5DEROD1cJ0NA
509
554
  sglang/test/simple_eval_math.py,sha256=6kGKNwNbLN-Af3Wj8WTimWhH-Xp3enDmSvvSjsgWUpk,2550
510
555
  sglang/test/simple_eval_mgsm.py,sha256=rd7TSUyxdKbrXaVoewo24V8lCo_6kO8zxPhhmvylpw8,10259
511
556
  sglang/test/simple_eval_mmlu.py,sha256=FkwamjGMjueTixymkedF-YiPloSLiy4ftILFUrKZ9XI,4357
512
- sglang/test/test_activation.py,sha256=jkdNRzJnbd5OgZliQaIXpxovlcky17UrweomcOcMxoE,1442
513
- sglang/test/test_block_fp8.py,sha256=bHYgQijsx0D4q0CgoZv6jOQLAWUz6QQC1cgT2oLLOnE,14420
514
- sglang/test/test_block_fp8_ep.py,sha256=hkuQjmCv3y_hWZj21cT9EaB6KSfT3JSzYPRQNFaLP-Q,10759
557
+ sglang/test/test_activation.py,sha256=GeTIJHxlLQfW3kM-X1FGa8Sa3dSGKHEXl5wEy-hfGis,1489
558
+ sglang/test/test_block_fp8.py,sha256=IqdQKt23annq_QR1gwVX0vzdMyWTEBLRhmPiLMemKI8,14458
559
+ sglang/test/test_block_fp8_ep.py,sha256=N1rvqbPErBaFFpeAw8TLYXGNZOoG7cfIBP2p5XbSyMo,10806
515
560
  sglang/test/test_custom_ops.py,sha256=XBTWh3jEsoe9hZ93p3LAUtjEj5l0qNEaZM5Mto6pbJA,3262
516
- sglang/test/test_layernorm.py,sha256=IacByD5d-stXjzBz8Ypamc7povlcedpKPbb_4JLgo3c,3720
561
+ sglang/test/test_dynamic_grad_mode.py,sha256=L76yUCuk_ymNpXD2CmO8r2GiGjIvD_gtTsuFDs2NolI,1638
562
+ sglang/test/test_layernorm.py,sha256=2GMWqqNDuGvSMSsEBF5eDCzwVSYA9E6hGhRo6s4ecKg,3764
517
563
  sglang/test/test_programs.py,sha256=VZ3vXtUDBnXz0M7gFdDH8hXg9Wa0j_qI8CVqjEgRN_E,18877
518
- sglang/test/test_utils.py,sha256=UBBk-F4A5hEtmRdu7vEttlnSmPVPJLQgtNZxMn1Tdo4,29188
519
- sglang-0.4.4.post1.dist-info/LICENSE,sha256=FJXh51fvTQklojUFY89XVLsjxRcBqOxPs8XNy-2uZ0c,11346
520
- sglang-0.4.4.post1.dist-info/METADATA,sha256=P0EmlGnVHDnvAojlrFQw_2CrPN2qERG6VIUoR6kG4vo,24311
521
- sglang-0.4.4.post1.dist-info/WHEEL,sha256=52BFRY2Up02UkjOa29eZOS2VxUrpPORXg1pkohGGUS8,91
522
- sglang-0.4.4.post1.dist-info/top_level.txt,sha256=yxhh3pYQkcnA7v3Bg889C2jZhvtJdEincysO7PEB09M,7
523
- sglang-0.4.4.post1.dist-info/RECORD,,
564
+ sglang/test/test_utils.py,sha256=bjRoX9pp98l9AKr0q9TgK8rVNxqMJ-1MeEZdrhSaMPw,30694
565
+ sglang/test/attention/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
566
+ sglang/test/attention/test_flashattn_backend.py,sha256=OxS1KsPs19nwZcDtdURj7_liT1cIfEXb6W4FH9KMaaE,10808
567
+ sglang-0.4.4.post3.dist-info/licenses/LICENSE,sha256=FJXh51fvTQklojUFY89XVLsjxRcBqOxPs8XNy-2uZ0c,11346
568
+ sglang-0.4.4.post3.dist-info/METADATA,sha256=jVtIt2_q-sSQu5V2AzV5Ex2pRWTJHW9PrNe_NuQ6WpM,25024
569
+ sglang-0.4.4.post3.dist-info/WHEEL,sha256=CmyFI0kx5cdEMTLiONQRbGQwjIoR1aIYB7eCAQ4KPJ0,91
570
+ sglang-0.4.4.post3.dist-info/top_level.txt,sha256=yxhh3pYQkcnA7v3Bg889C2jZhvtJdEincysO7PEB09M,7
571
+ sglang-0.4.4.post3.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