sglang 0.3.6.post2__py3-none-any.whl → 0.4.0__py3-none-any.whl

sglang/test/test_utils.py

@@ -22,7 +22,7 @@ from sglang.bench_serving import run_benchmark
 from sglang.global_config import global_config
 from sglang.lang.backend.openai import OpenAI
 from sglang.lang.backend.runtime_endpoint import RuntimeEndpoint
-from sglang.srt.utils import get_bool_env_var, kill_child_process
+from sglang.srt.utils import get_bool_env_var, kill_process_tree
 from sglang.test.run_eval import run_eval
 from sglang.utils import get_exception_traceback
 
@@ -424,6 +424,7 @@ def popen_launch_server(
         port,
         *other_args,
     ]
+
     if api_key:
         command += ["--api-key", api_key]
 
@@ -504,7 +505,7 @@ def run_unittest_files(files: List[str], timeout_per_file: float):
             )
             assert ret_code == 0
         except TimeoutError:
-            kill_child_process(process.pid, include_self=True)
+            kill_process_tree(process.pid)
             time.sleep(5)
             print(
                 f"\nTimeout after {timeout_per_file} seconds when running {filename}\n",
@@ -567,6 +568,7 @@ def run_bench_serving(
         disable_tqdm=False,
         disable_stream=disable_stream,
         disable_ignore_eos=False,
+        lora_name=None,
         extra_request_body=None,
         profile=None,
     )
@@ -578,7 +580,7 @@ def run_bench_serving(
             run_benchmark(warmup_args)
         res = run_benchmark(args)
     finally:
-        kill_child_process(process.pid, include_self=True)
+        kill_process_tree(process.pid)
 
     assert res["completed"] == num_prompts
     return res
@@ -611,7 +613,7 @@ def run_bench_one_batch(model, other_args):
         lastline = output.split("\n")[-3]
         output_throughput = float(lastline.split(" ")[-2])
     finally:
-        kill_child_process(process.pid, include_self=True)
+        kill_process_tree(process.pid)
 
     return output_throughput
 
@@ -677,8 +679,14 @@ def run_and_check_memory_leak(
     enable_mixed_chunk,
     disable_overlap,
     chunked_prefill_size,
+    assert_has_abort,
 ):
-    other_args = ["--chunked-prefill-size", str(chunked_prefill_size)]
+    other_args = [
+        "--chunked-prefill-size",
+        str(chunked_prefill_size),
+        "--log-level",
+        "debug",
+    ]
     if disable_radix_cache:
         other_args += ["--disable-radix-cache"]
     if enable_mixed_chunk:
@@ -710,8 +718,8 @@ def run_and_check_memory_leak(
     workload_func(base_url, model)
 
     # Clean up everything
-    kill_child_process(process.pid, include_self=True)
-    kill_child_process(process.pid, include_self=True)
+    kill_process_tree(process.pid)
+    kill_process_tree(process.pid)
     stdout.close()
     stderr.close()
     if os.path.exists(STDOUT_FILENAME):
@@ -723,14 +731,19 @@ def run_and_check_memory_leak(
     # Assert success
     has_new_server = False
     has_leak = False
+    has_abort = False
     for line in output_lines:
         if "The server is fired" in line:
             has_new_server = True
         if "leak" in line:
             has_leak = True
+        if "Abort" in line:
+            has_abort = True
 
     assert has_new_server
     assert not has_leak
+    if assert_has_abort:
+        assert has_abort
 
 
 def run_mmlu_test(
@@ -761,6 +774,7 @@ def run_mmlu_test(
         enable_mixed_chunk,
         disable_overlap,
         chunked_prefill_size,
+        assert_has_abort=False,
     )
 
 
@@ -800,4 +814,10 @@ def run_mulit_request_test(
         enable_mixed_chunk,
         enable_overlap,
         chunked_prefill_size,
+        assert_has_abort=False,
     )
+
+
+def write_github_step_summary(content):
+    with open(os.environ["GITHUB_STEP_SUMMARY"], "a") as f:
+        f.write(content)

sglang/utils.py

@@ -348,9 +348,9 @@ def wait_for_server(base_url: str, timeout: int = None) -> None:
 
 
 def terminate_process(process):
-    from sglang.srt.utils import kill_child_process
+    from sglang.srt.utils import kill_process_tree
 
-    kill_child_process(process.pid, include_self=True)
+    kill_process_tree(process.pid)
 
 
 def print_highlight(html_content: str):

sglang/version.py

@@ -1 +1 @@
-__version__ = "0.3.6.post2"
+__version__ = "0.4.0"

{sglang-0.3.6.post2.dist-info → sglang-0.4.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: sglang
-Version: 0.3.6.post2
+Version: 0.4.0
 Summary: SGLang is yet another fast serving framework for large language models and vision language models.
 License:                                  Apache License
                                    Version 2.0, January 2004
@@ -241,6 +241,7 @@ Requires-Dist: sglang[runtime_common]; extra == "srt"
 Requires-Dist: torch; extra == "srt"
 Requires-Dist: vllm>=0.6.3.post1; extra == "srt"
 Requires-Dist: cuda-python; extra == "srt"
+Requires-Dist: flashinfer>=0.1.6; extra == "srt"
 Provides-Extra: srt-hip
 Requires-Dist: sglang[runtime_common]; extra == "srt-hip"
 Requires-Dist: torch; extra == "srt-hip"

sglang-0.4.0.dist-info/RECORD

@@ -0,0 +1,184 @@
+sglang/__init__.py,sha256=3M0oz0ZA8fULhV5LwQ4hxh-MRdHsOJRD1D63C60pdG4,1616
+sglang/api.py,sha256=NdO6cYnklnEBQBKqQjlqI8-P1EownKQ71t5ibCGhEVo,6953
+sglang/bench_latency.py,sha256=oZjSAzX7dUiSu-zdz0dkyUPo-qAX_lsXFH1gf03akgI,76
+sglang/bench_offline_throughput.py,sha256=3OrFI26PmoVTU3pQrBFC50AZI7HpKKuk4vYycbkDjhY,12428
+sglang/bench_one_batch.py,sha256=vxXSCQRTMeJUtJKsSoP6tLdoWTdFp1mhwsLpKHccs2c,15858
+sglang/bench_one_batch_server.py,sha256=-fV9FTLNNcSIy0pgYeggXedPVK0fVsXZqVQswT8OMOY,5945
+sglang/bench_serving.py,sha256=Oa_Qi7YApv37jGDAmuIaZSIhayvRpKq9GZGZLXBU-9I,52924
+sglang/check_env.py,sha256=q1sdYL-gcKSCeIZMk7sUMh9rjM71f-EUgp07OGPSbZM,5446
+sglang/global_config.py,sha256=fnT0U9vlHdGaQFKN9tYTnUF4-eVW4HYQURd5zvPtrg0,1286
+sglang/launch_server.py,sha256=4y2QeSj0wVNB9MJQZeahD4ahTDU6gwqo7MPUytyFop0,403
+sglang/launch_server_llavavid.py,sha256=tGc17S1vUfLwbi1GB26oOdXxTWr7gjlqpTrPnrMRNO8,1007
+sglang/utils.py,sha256=r4Dw-xffcrTRposls-gqyoYxjgJNYhVduK_6bDN_Vj4,11526
+sglang/version.py,sha256=42STGor_9nKYXumfeV5tiyD_M8VdcddX7CEexmibPBk,22
+sglang/lang/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+sglang/lang/chat_template.py,sha256=jprS3-In2FTUoedKwZg-HYvDwU8RTIYntOlf2zoN2sU,14814
+sglang/lang/choices.py,sha256=-W1DVw9N9ZliVpvmWrzIXG4cswAah8eMQrHWzkS3D8o,6234
+sglang/lang/compiler.py,sha256=o1C6G3TzhjSlsH-doTPy5oiVehr57dxNTa5oZw5TTAI,7639
+sglang/lang/interpreter.py,sha256=SBjejhLhTKzNM0HbjtTg5r17WPJ64WFSk6lcM_SCWKs,30717
+sglang/lang/ir.py,sha256=zpzzAO1YVldhE95Vwz5hU_TQltu-xt8A6rfFr0PuIDA,18410
+sglang/lang/tracer.py,sha256=o-jLAPPSuy2vBfsGGrTAnbuWtORzQ50B4C_P5zvYkx8,8291
+sglang/lang/backend/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+sglang/lang/backend/anthropic.py,sha256=EXRX7xJgA5KZszX7toSLVnKzFQ5EO0Loj-YjHFtxSxg,2081
+sglang/lang/backend/base_backend.py,sha256=tdoh9YF3CyekY1BKiX9n7-aA4srDWIuA4RDJLM7q8qg,1985
+sglang/lang/backend/litellm.py,sha256=ugmL7sfUxkUHVbHtwNzHgdQAEd4UCjNQboFuE3KThcY,2450
+sglang/lang/backend/openai.py,sha256=qM7eVH_kMxnDd2rpxOH0v76KxtOJFlAwgLgWIKvFGCI,15060
+sglang/lang/backend/runtime_endpoint.py,sha256=IWbrAKrUkzNOvwV6V9_y6pkTr2SUYEkKBT-3kirgad0,10514
+sglang/lang/backend/vertexai.py,sha256=O-iBLD-y3vq80UxnrAoJri7bxpgd-_eakZ88Cf8bEGA,4855
+sglang/srt/_custom_ops.py,sha256=Y4gyTDGhWz-W2Igq25Ojm8XFiyvkawW9I-79iwYvxJ0,3574
+sglang/srt/conversation.py,sha256=u9zFU8aMYzwHUbQRKU76B_T-jfLlPoxUcWG_nRbDM2I,21201
+sglang/srt/hf_transformers_utils.py,sha256=38Ms0H2-VMerOS6jnczcFtZMS6lhw9B5rSWKAfxVUfQ,7945
+sglang/srt/mm_utils.py,sha256=1ScBunw_x4W8ebM_AcJ62-1T2mfT8NlMJqdAhkF1lb0,12367
+sglang/srt/model_parallel.py,sha256=QR-Alqo0sElDXPJ79N1PhUHHKiEHPQn3dyXduMP-SHQ,3664
+sglang/srt/server.py,sha256=uApz59VUGkRx_HCy6kaiKJFTmaOAvAM3o2yYjEWi9EM,34594
+sglang/srt/server_args.py,sha256=NubTM6ocZY0YDSvPQbvBdURK-7E5kSKHA2Ze09O8_1I,32182
+sglang/srt/utils.py,sha256=p-eYNMVQkw6Yw1b7MBSNWhVvykCuNfejqMjl7cLDq7A,40708
+sglang/srt/configs/__init__.py,sha256=_usVIXHQjft4PAJ1Y-yGQOn2QNOv501GYMlQwpGXbns,208
+sglang/srt/configs/device_config.py,sha256=dResqHjkg_dq10v6rnVpbXpvABZRB0jylOm-2_JAnx0,428
+sglang/srt/configs/exaone.py,sha256=Duxd4yQoKy8GWEzZD_kCY_OzmN_67CTJL_Kgn0eXk3g,10731
+sglang/srt/configs/load_config.py,sha256=TcPi_HY6xu5SiVZsxPOoB5pGeDUNebOk7muoUH9VBDg,3083
+sglang/srt/configs/model_config.py,sha256=OjEeigs5tMNKP-RImJk2NHVFXv-fyQfsGREWMO3rqhM,15839
+sglang/srt/configs/qwen2vl.py,sha256=ZjLy9v2eZY4wptUfY3CWgYKg2B5DDrkfCSyTy_Zf_bg,4351
+sglang/srt/constrained/__init__.py,sha256=UWZNVLvOT5ZBX8M36sONgDmnKtkQ0cSfhQD2jO0ATuk,786
+sglang/srt/constrained/base_grammar_backend.py,sha256=FhVm7PxhXDl0joV9NP5RjKgz7dR1dZvUAQnh0mdtvVY,2353
+sglang/srt/constrained/outlines_backend.py,sha256=LmezsMAmTlQgaPqO4Axl4EcSAqKr_ZYZASfAoVxT17A,6670
+sglang/srt/constrained/outlines_jump_forward.py,sha256=iZWXeR3gNYoMubLGyFmLPO4V2YsN5DiGjD71Xk9iFaE,6418
+sglang/srt/constrained/xgrammar_backend.py,sha256=4ZCQgcjWEY2Lg4r2V9sAiYJJblkQ_uVbEnvsjqhR1Pc,4548
+sglang/srt/distributed/__init__.py,sha256=__tl9Frrf3PFrSyNYcn5i-y2rL-J4-Qn6RJwrsZ4xgc,83
+sglang/srt/distributed/communication_op.py,sha256=ZoIhboZyefiAwr-1K-wF3rAFSQ4Wt-RxXpsX443Gbt4,1157
+sglang/srt/distributed/parallel_state.py,sha256=HplRH5S0AWdwSdhoHYX9_UWQZlFjh2Z1LHaz68EXlpE,47555
+sglang/srt/distributed/utils.py,sha256=riYflM9l1-Yi-8Ce8Acxa4mAjZaxHRQfta8Dtah4yG0,8500
+sglang/srt/distributed/device_communicators/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+sglang/srt/distributed/device_communicators/cuda_wrapper.py,sha256=u8_kYVOBDrNZMiQCJC538yJvpZgq6ZEpB28tCrp04yM,7065
+sglang/srt/distributed/device_communicators/custom_all_reduce.py,sha256=5ARfr-1_V4QoxjvdfxOKPtSK_Rax8qAQTPoA5z_Emtc,13567
+sglang/srt/distributed/device_communicators/custom_all_reduce_utils.py,sha256=qq8GTZl0br0ggfosb8mH3U6cXbm4NWfr8y_B83W4fDg,11081
+sglang/srt/distributed/device_communicators/hpu_communicator.py,sha256=LXGOhoNT5iVu1JWlRvGfHMB0wRW6lkhDamVT9JhVD94,1755
+sglang/srt/distributed/device_communicators/pynccl.py,sha256=cDEoHU24C8ph-4fJAIDjZfl53aSzrjCG3FAFkt4vjrM,7186
+sglang/srt/distributed/device_communicators/pynccl_wrapper.py,sha256=pU4xhG-WKytSHJ-cpcPEs0WG4dAg44jpOgv2dAmHisE,11990
+sglang/srt/distributed/device_communicators/shm_broadcast.py,sha256=WVxBd1QfIgRWzVGtN2axxO-3PFT-Qww8GQ82Yg5PPYU,22824
+sglang/srt/distributed/device_communicators/xpu_communicator.py,sha256=P3WKgddcfpUhBa-_5PvjYxH146ZE-N1cotTzEpPRKlY,1620
+sglang/srt/layers/activation.py,sha256=EboMjT9HV2tNHQ6rzpojtlkzev1lAFbhQlxMg9hwxBQ,5471
+sglang/srt/layers/custom_op_util.py,sha256=0vu-yX2wwonmO1L_o5G7SA6C-8XuhDIh9rPDvNeLhoc,922
+sglang/srt/layers/fused_moe_patch.py,sha256=DMIyrwOON7OSidKZdreL5HzMhP0AD5Ues0xdY-ADOQw,4471
+sglang/srt/layers/layernorm.py,sha256=nRQ1w1xSUcU-zlqVC61BnGG6otS5W1w9VaSzeXizrx4,4037
+sglang/srt/layers/linear.py,sha256=dF2HvqiMbhWlCjvkLFRCcgUFGhG-B0keM_CIpjvgTtg,46154
+sglang/srt/layers/logits_processor.py,sha256=oZNu9pNNgmswhuw8irlLm0SfpVrD7cFf-GdfPsLZGHE,13227
+sglang/srt/layers/pooler.py,sha256=rj2lygvleBnyLCBZ8I11HGMgpfIDsT0l3PIkshJwdu4,1606
+sglang/srt/layers/radix_attention.py,sha256=C_mK4mfmKlxMRNeKYP9E5R3PRd3eT-OcE_g3mo36dJM,2058
+sglang/srt/layers/rotary_embedding.py,sha256=29tx3JNR40AoXqBa2cFGBjva9vU2xgFipETlpMaaZas,3985
+sglang/srt/layers/sampler.py,sha256=_enfER8MSxsCYrR6_NgyFxKA_XqKtii_asOZUFUUsd8,4580
+sglang/srt/layers/torchao_utils.py,sha256=v0hyr4hLsM42QwOPCdKb-ftRTjVokBZbqvRj4O4C-Nw,3415
+sglang/srt/layers/vocab_parallel_embedding.py,sha256=slGwLiWjuFLCUdRe-GTlfumyZpqVX9VF6No_UGOT-hA,21624
+sglang/srt/layers/attention/__init__.py,sha256=EL1o6Q5vLgViN3pOr2A7F6K9FlNEpMdBypFAVMeq_HA,2445
+sglang/srt/layers/attention/double_sparsity_backend.py,sha256=BlX7uXteQpnoOnKsdBKh8h20zMVMEiibB5F_PkZSlNI,10706
+sglang/srt/layers/attention/flashinfer_backend.py,sha256=q0yPeKUjGbxKnOqbZLHs5fyYkkVE3YEkBBMWtUaiDL4,24611
+sglang/srt/layers/attention/torch_native_backend.py,sha256=amR8xGaaCONvo3M8I4nuYMv6OxPHyYoxu3m8vAqjWfo,10295
+sglang/srt/layers/attention/triton_backend.py,sha256=gjxed2cvc2-8QEHkzyTVv6ui7oYOp2b_vgIUQVD1XuM,6538
+sglang/srt/layers/attention/triton_ops/decode_attention.py,sha256=BE63WhKiutSNkhJLsRwvfsRy-ExvuAv7FZyoWv73ul8,18744
+sglang/srt/layers/attention/triton_ops/double_sparsity_attention.py,sha256=1pSXfY3EEaM7iRN_uElHnAfsrJMhTFbu9fj8Z0O2PbE,21480
+sglang/srt/layers/attention/triton_ops/extend_attention.py,sha256=Gfct-0_l-S2ZrP4F-zkzNiFbmd3C3f7uJovacOuDxaA,11472
+sglang/srt/layers/attention/triton_ops/prefill_attention.py,sha256=lojFXRZMLWkzS2Y8uxaolnQhXaWKG19mCAWaF5KQeiI,6087
+sglang/srt/layers/fused_moe_triton/__init__.py,sha256=PHKFqd2hPOO-g9kSMseg2g76lpg9OGXQDThWU6bt9vs,902
+sglang/srt/layers/fused_moe_triton/fused_moe.py,sha256=qwfRBOeY5DT48Q6z71Eh9cjFehvs_K6eLIVWNL044Ug,28363
+sglang/srt/layers/fused_moe_triton/layer.py,sha256=URDkTt8xEqnqpO5tb_3L7JlhlO53VWfqDDNSRYEu-LY,21545
+sglang/srt/layers/quantization/__init__.py,sha256=0LnBJ2LOtk7HnMzOZm14tMd4Y0eSPSgqUhj9uOn53Es,5961
+sglang/srt/layers/quantization/base_config.py,sha256=daK9p0aijMszLUm1W4Pc33FK87MdqYK1NoWFKif-j80,4599
+sglang/srt/lora/lora.py,sha256=-o2mBmUvoVpdkgdAkWTARN4kfyep3UNEJLcg6moh0SU,15056
+sglang/srt/lora/lora_config.py,sha256=a2fTQESlCbG1xLiBYy4ptZ6c0Burcqyg1_6V1XSok-Y,1506
+sglang/srt/lora/lora_manager.py,sha256=DHiqdl0_4wQ5PxZBZtlCpP14515mDV2_H9tzL3Rdss8,12886
+sglang/srt/managers/data_parallel_controller.py,sha256=psI4FAuBGjtdnEuwagnGdtRqvqSSxOROfNKQqVDqlVA,8382
+sglang/srt/managers/detokenizer_manager.py,sha256=TtrtE37XT5XcJzk8-R5rHZ16NHTPd5XZi8hf3h-sB2A,7462
+sglang/srt/managers/image_processor.py,sha256=Y8RgyrzbJjJTpjbnZDa5qiiG5wWjZ68rOXUPDi6kkFo,13698
+sglang/srt/managers/io_struct.py,sha256=d_kctmHcNBzzaP5lUEIpdrVVsob4dNOetMHkobUJZz4,14439
+sglang/srt/managers/schedule_batch.py,sha256=p6OmeCH3arGZ99Ch7JrnTgGU-SsPgjQY7Z6arVzeqao,44998
+sglang/srt/managers/schedule_policy.py,sha256=7QuIsJDRzkrvs3IJk10oOfL4Me0UZwDYvRniT1fSFuo,12620
+sglang/srt/managers/scheduler.py,sha256=fmvwVcZS5DsDdxIJ4bjlLV-qJsRbTxjsqPP5a2SS0C8,59372
+sglang/srt/managers/session_controller.py,sha256=Yp-IV3rXczACZxZXmF-QxW9CWICGy8KHQ9ttBGJ8WXA,2800
+sglang/srt/managers/tokenizer_manager.py,sha256=XPaSXB6b23u95viFqlqd-tdyrNMMOOiSDWviz_g7UBM,29890
+sglang/srt/managers/tp_worker.py,sha256=X1EwFX3FSsmXx7jeeX2tjZRocaujabQYWm-M-0CFEBE,7363
+sglang/srt/managers/tp_worker_overlap_thread.py,sha256=ZBLCAYz-Ls1coObyO8dNtNsO6gG2rn40KulEmALB_J8,8686
+sglang/srt/mem_cache/base_prefix_cache.py,sha256=qEQwEkG4E5rab2ZoTqcesf5pR_J4nV2jBxIHsBJHtIM,924
+sglang/srt/mem_cache/chunk_cache.py,sha256=VcCpyrf5FOQ5xoKeOouCI5ZQLkZo_pgY1SPbDDkagGg,2492
+sglang/srt/mem_cache/flush_cache.py,sha256=GYcxmNXh4hsMpFfNOuCTpKilW7guZwTtAg_usVeM3J0,979
+sglang/srt/mem_cache/memory_pool.py,sha256=41fjuj_sD0yfJq-sy-X99cc2djBa6w4dy2y47V0WqNU,10934
+sglang/srt/mem_cache/radix_cache.py,sha256=DzLCO_gYQ7X_C2NJSEHzzMZhb5HzWjKF9wXJQsnzr8M,10427
+sglang/srt/metrics/collector.py,sha256=ZWoFx_FKN0sNMSZ8RJWUVQ0RFEYhIHxdw0d4TZTluMU,6861
+sglang/srt/metrics/func_timer.py,sha256=VFyNRrbnKVCwnQsrlLin1lITJfjQpf9m8sGPqL5LIsQ,3438
+sglang/srt/model_executor/cuda_graph_runner.py,sha256=ffypnhzwAdiBAngHEpK9HVmpvX0H3dFIKGa45tS0IGI,15197
+sglang/srt/model_executor/forward_batch_info.py,sha256=L5mVoW5SaO6To-7nGk0TZM-FFB5_78cARpJ-aC2rwD0,12883
+sglang/srt/model_executor/model_runner.py,sha256=wU06s5tpHgtxitmAWGQC1NjLIcp68iAdbFDWN-TtkBU,29260
+sglang/srt/model_loader/__init__.py,sha256=zGZkOBz1zx-pkaIy47BasL3fjDlAcxAXUTjInOhXHAE,919
+sglang/srt/model_loader/loader.py,sha256=VBrY4W9CiVvS_D8yXhdkW9jReV9rSMSkJplabz0Fxgk,43528
+sglang/srt/model_loader/utils.py,sha256=0NaMR67fESFopaklmsleiL27XH1QUrjZW246MUu1EJ0,1369
+sglang/srt/model_loader/weight_utils.py,sha256=kQo9KPThjH3HAOCfC_tdwdrshdWuWJOVpPR0skSyaRY,24193
+sglang/srt/models/baichuan.py,sha256=PzBOFcEAixakPEkQSaJwC0Xc1fu-yCsN9T0I67r8QmY,14919
+sglang/srt/models/chatglm.py,sha256=DOrEhmb0s-yPId88R6nJeLOTUEtogk-vkB69qT2JdWc,12913
+sglang/srt/models/commandr.py,sha256=-cswAK2_ZTcOsvBkv5Z8SZ1x60cxJryslbSPHDOcnbM,14169
+sglang/srt/models/dbrx.py,sha256=2Wqcf3sv57l4gi2xH8yrb5WSmY-4_kbbf6fhpJ4aKWw,14581
+sglang/srt/models/deepseek.py,sha256=BVNICGoLjQoHmR5lc31YrZ6YbxSRTBilHqlLsALr2u8,15693
+sglang/srt/models/deepseek_v2.py,sha256=2AWdDb5qZ6dj_M6ZY-rCG0omTmerUxbhwuvI9Um4Bg4,31967
+sglang/srt/models/exaone.py,sha256=dkERTZVxrRroqu5AGLP7D4N6n8HvDqlNaDQUIe15mZY,13038
+sglang/srt/models/gemma.py,sha256=ydRqsG-7004r1fAiz01LHUmcj_6XN0Tn4xO1keJnMQk,12126
+sglang/srt/models/gemma2.py,sha256=vPrAasJajitQHB9ZqMFut58xNsOm3fk2m05a-feQL10,14600
+sglang/srt/models/gemma2_reward.py,sha256=hJw0hXNPyQSpazkVJVYiW04OtTZH0GiLI-JJef_kaGs,2529
+sglang/srt/models/gpt2.py,sha256=2je1kE09sGcaORWnJuGYAkcwwOrT9EK-KhQaoCKjCSA,9517
+sglang/srt/models/gpt_bigcode.py,sha256=tovyOdJu2x3LkzmkdFXX_iJdkxuyChIDxwgvPBy6UPo,9528
+sglang/srt/models/grok.py,sha256=lJQEk2D8zyAXE0g4vXLCEmiWM938K_qly02EScwwV_k,13942
+sglang/srt/models/internlm2.py,sha256=_xcKtd6YtEFUTozaN-yUb0xbSYckRpomfPSKcAk4j-Y,12127
+sglang/srt/models/internlm2_reward.py,sha256=8K26A9oIFFGx_9U2mF87j7FX8K87HGKMnVL3ht1Uc7I,2398
+sglang/srt/models/llama.py,sha256=JVwVDU-8L8RSNOYH0CK8FI557xqjPU2Ts_pQhVfGZv4,19550
+sglang/srt/models/llama_classification.py,sha256=EdXmiMyfJ9NH5P-Wel7SRhf_v8ddFFhVJMQgzDt0oVk,3377
+sglang/srt/models/llama_embedding.py,sha256=rh-AiczPY_pTpzcACHvSMVjh1hsV_MZBBwP0LQxPsGM,3130
+sglang/srt/models/llama_reward.py,sha256=JVaiTK4gVXNMimeq3kKkv7dt5Hc77hPqF4ewvmzjJes,4622
+sglang/srt/models/llava.py,sha256=l9mqS9wl_l6ARC-K1UUe7XsB5k9sZratMNQEwx5IjR0,25229
+sglang/srt/models/llavavid.py,sha256=dYUkKfHoE15vF_VXA_s_ICCTUMSmSgvP181fk8dUi0g,12185
+sglang/srt/models/minicpm.py,sha256=ws4AqhOfAvYHGd04QuXCZel-Oxy9_vN4p4rTjs9RSz0,13723
+sglang/srt/models/minicpm3.py,sha256=YIKJDTpwjmpLlv1sNT93k2yZMvGQlI_H87czjf6QYyo,24707
+sglang/srt/models/mistral.py,sha256=EYifJUUzN2Z2-iL37eJiNZF_DB0H4pa0mKlgYRIxM70,838
+sglang/srt/models/mixtral.py,sha256=KvvtLWn-GLtwbdrt4PBq9gIFJiJH-JFcj0BLfcLcXjo,14322
+sglang/srt/models/mixtral_quant.py,sha256=uuVO1nWUZJiDhbqZN6gzSMwyfpyZorMuFXHeMCGo7N0,14022
+sglang/srt/models/mllama.py,sha256=3kX-UqeTSYZL5kPNdkfKEAEv3DpSAW1ArAAoeiXVzIc,37739
+sglang/srt/models/olmo.py,sha256=OCDMtX1OI83r80mzU4FMC3Tg8cleQ-7C8Tpoe8zgzss,11708
+sglang/srt/models/olmo2.py,sha256=aC7svioN7XT5owRxPrvhvWBNMON9QXGQBWJ1KHMyXeA,13442
+sglang/srt/models/olmoe.py,sha256=Rw-3YrHWd90MZQFnmcfUQ-3wAaI0PCFKb0DIrCDND3s,15347
+sglang/srt/models/phi3_small.py,sha256=rQFv-4aUGkNPTQjTEME5bcOL9DEQqqFb3aDKrj2joOA,15083
+sglang/srt/models/qwen.py,sha256=_FKDbwaS5C07uJyyivZpBrXJVej4Ph9ivzJdzWJPxJ4,9904
+sglang/srt/models/qwen2.py,sha256=Kh6mW0H2jQdrPS9dJnJShLpo0BNEq6oI4oy5VMHGzac,12444
+sglang/srt/models/qwen2_moe.py,sha256=5-ZTYW5bERhTG5LnNXMqYfWB17NBkvxmQJu6DuDAFCo,16819
+sglang/srt/models/qwen2_vl.py,sha256=3EaUlTbyWOTRXA7eViK1WqmVbCFhXLIpnos49zzf-yM,26561
+sglang/srt/models/registry.py,sha256=inKh9iwOp3LFYm3nqujg-OtABClOP-ifc1stA9cZegA,3434
+sglang/srt/models/stablelm.py,sha256=iBlIkM7CQmqI25nsujWk0LLCQD7TshzUU8qzZYYrt20,11311
+sglang/srt/models/torch_native_llama.py,sha256=dHuUln7RcsRcdyOIUw7HKtOOG9OPmnfOpK8j5gYXzJY,19039
+sglang/srt/models/xverse.py,sha256=Oq--KqvbYu2H4TMVGEHpSnJLEwXBpxlncR9ilsQeckc,13579
+sglang/srt/models/xverse_moe.py,sha256=AawKEQw--oAl-yzwCjoaZRG7q3rdkyDiam3FS0zjf_c,15537
+sglang/srt/models/yivl.py,sha256=88OubtuZ38Dxb2LzfV_MTPBI4wKhh4NJqFu--efbhFM,4809
+sglang/srt/openai_api/adapter.py,sha256=gZEaG1dVSFv9WLj0369Ke1yrNNgi_gpgKxPt5Ju9mUw,53775
+sglang/srt/openai_api/protocol.py,sha256=4T9hGCrpfCUSBjKZFvemTfj49CkTUzpCcx6izLv3ir0,10246
+sglang/srt/sampling/sampling_batch_info.py,sha256=YC-KPyDWyLGNPL4YVcst4xwP8Wlz2zcCNJHB_5zljXQ,8470
+sglang/srt/sampling/sampling_params.py,sha256=n7RbBg_bS5fYhsiWa8uJYnfoXy_i5DvtTBOkuFnHDNU,5286
+sglang/srt/sampling/penaltylib/__init__.py,sha256=5vQw0Y5DSzmsoFg1IdMIKLwFVhYZ5ArADHVBYbSmOec,513
+sglang/srt/sampling/penaltylib/orchestrator.py,sha256=J-DEemZcKm1--o37kf3qDOE8SZ_6H3d5oex49Mgq2ZU,10762
+sglang/srt/sampling/penaltylib/penalizers/frequency_penalty.py,sha256=1Zp2aL6dD60mwD1tCcSG0x5IYo0v4z9ce-q_YwbJ9f8,2490
+sglang/srt/sampling/penaltylib/penalizers/min_new_tokens.py,sha256=_Nxv0XgUPirZjw2SEJYp_Cd9ZcLwmt7h6JE6J4hhFq4,3629
+sglang/srt/sampling/penaltylib/penalizers/presence_penalty.py,sha256=5tOgCg7OvE9kSN9VMCpH1hwqo1YMxt9iS5PVpct9HpU,2468
+sglang/srt/sampling/penaltylib/penalizers/repetition_penalty.py,sha256=m22Rfn1RuB1HpImBDECsiJ2VooBYpsFADAwnk1EPzk0,2751
+sglang/test/few_shot_gsm8k.py,sha256=7yDbEQe49gZeJhz2wFFX-gf_59ThDKsCS1xwfogNc7k,4034
+sglang/test/few_shot_gsm8k_engine.py,sha256=QQbrwOX6-cJDD3RZC_e7zPnt6aSo8JdF8X_lRHSjdDM,3886
+sglang/test/run_eval.py,sha256=9yO0hXZOcn4abEOs96T-XPguDEklK16Ltco0pGF3zCg,4020
+sglang/test/runners.py,sha256=ANzjrHkT_1E0G3UcD47O8XEKst3Si4AOfx-uErbFS7o,15129
+sglang/test/simple_eval_common.py,sha256=joqrGysuLnJFtzDRIgFkMsRyKUSyjVPFWp0_PHAL3Ik,12378
+sglang/test/simple_eval_gpqa.py,sha256=8Xt9Bw05c7SZTYrCZgB68OZUqUbLo69ywiyx0bTvSUk,3220
+sglang/test/simple_eval_humaneval.py,sha256=zmV3xWYc2OrpiT9Dy55RTKZL5DEROD1cJ0NA_-cU5zI,5685
+sglang/test/simple_eval_math.py,sha256=6kGKNwNbLN-Af3Wj8WTimWhH-Xp3enDmSvvSjsgWUpk,2550
+sglang/test/simple_eval_mgsm.py,sha256=rd7TSUyxdKbrXaVoewo24V8lCo_6kO8zxPhhmvylpw8,10259
+sglang/test/simple_eval_mmlu.py,sha256=FkwamjGMjueTixymkedF-YiPloSLiy4ftILFUrKZ9XI,4357
+sglang/test/test_activation.py,sha256=jkdNRzJnbd5OgZliQaIXpxovlcky17UrweomcOcMxoE,1442
+sglang/test/test_layernorm.py,sha256=IacByD5d-stXjzBz8Ypamc7povlcedpKPbb_4JLgo3c,3720
+sglang/test/test_programs.py,sha256=1Z0umrsUu9pagzyGH5SrXl_qhKSyTfUv_kWC2mcn0qo,18208
+sglang/test/test_utils.py,sha256=0lY3ZNfS3JCB4LqSRJgBfB8I0MA8TUT-BJmnrvQC8vw,23797
+sglang/test/srt/sampling/penaltylib/utils.py,sha256=CjxHgywh0hx_87iynzQt_ztHu6zBVuE-YrZ-XPmW6U4,12906
+sglang-0.4.0.dist-info/LICENSE,sha256=FJXh51fvTQklojUFY89XVLsjxRcBqOxPs8XNy-2uZ0c,11346
+sglang-0.4.0.dist-info/METADATA,sha256=dXh43Qx9ImYKIF-eB8vuiz_0DDbmNgUrOjCukOXCBTA,22165
+sglang-0.4.0.dist-info/WHEEL,sha256=PZUExdf71Ui_so67QXpySuHtCi3-J3wvF4ORK6k_S8U,91
+sglang-0.4.0.dist-info/top_level.txt,sha256=yxhh3pYQkcnA7v3Bg889C2jZhvtJdEincysO7PEB09M,7
+sglang-0.4.0.dist-info/RECORD,,

sglang/srt/layers/fused_moe_grok/__init__.py

@@ -1 +0,0 @@
-from sglang.srt.layers.fused_moe_grok.layer import FusedMoE, FusedMoEMethodBase