tpu-inference 0.11.1.dev202511150811__py3-none-any.whl → 0.11.1.dev202512030818__py3-none-any.whl

{tpu_inference-0.11.1.dev202511150811.dist-info → tpu_inference-0.11.1.dev202512030818.dist-info}/RECORD

@@ -1,9 +1,9 @@
 tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tests/test_base.py,sha256=Ct5WFRMHL7IHEIxk8FrzAvO8m0xFuDpzDBKkAKKAL2Q,7341
-tests/test_envs.py,sha256=Woyfp_d5HS-uTGo4_u9dYlBbgmhfIEoFb-Rx_k7YXD4,6298
+tests/test_envs.py,sha256=h502VxL2gvhECm8u5uDh5JTGvhFf_DfQO88SpqOFMzE,7135
 tests/test_quantization.py,sha256=IT5ASyS1uuWcxc22kRtBcA-V4j3Z3hb7pMztm3GOlBs,34445
 tests/test_tpu_info.py,sha256=ZrwlMsp8ffITkS_b8Q1t_QG-a-WVAd4NUcjHhGibcsI,4670
-tests/test_utils.py,sha256=Mta5ZzYCgRAh1-BjcOvvx9iQ9DnnXLps7oDHxVQp2yE,8236
+tests/test_utils.py,sha256=GIXLdd-x4gnqSLrySXGk22phqPc8MegFd7ph1Jj8OcU,8182
 tests/core/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tests/core/test_core_tpu.py,sha256=r496rk1eOsK_F4nvm9zprl_T-RcO6eCUb7LuVReOZno,21413
 tests/core/test_disagg_executor.py,sha256=QdE2YZs08EyDDCmSjhiXkXqQ9BJTgO6csr_E1xkkfSg,2256
@@ -11,7 +11,7 @@ tests/core/test_disagg_utils.py,sha256=alktTGppaGdg-_un0Amz8Y0IDQz-xNJN0dXG-YApE
 tests/core/test_dp_scheduler.py,sha256=IwCR1Vs96V4CQdWA051rNaYxxr2V_byA1yx9HWyRoMg,37339
 tests/core/test_init.py,sha256=NEFI5A9eKGu4rmeJ2iqd0EmhlA3bzbVkXmMi1PV1b9U,1687
 tests/kernels/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tests/kernels/fused_moe_v1_test.py,sha256=c6zbSHQDzOseeyL9VCjQeP7zayNnwYf059CPlKcvZzQ,3137
+tests/kernels/fused_moe_v1_test.py,sha256=sQ6gvpti94fpPYrSZn7frPPNjqbVmRibFtenVrGGCA4,10403
 tests/kernels/mla_v1_test.py,sha256=oZc4TCgquiG0KOeWfv46yJbUIpro_CgCMFc7vzyB7t8,11646
 tests/kernels/quantized_matmul_kernel_test.py,sha256=od5-zXFjcsc_gWGRDrREL8E_ftymNniQVTzgtkBo_Gc,5679
 tests/kernels/ragged_kv_cache_update_v2_test.py,sha256=6-HjP5CoUG-kcuP8MS-JJVMiBnPRo_zadS3VInnO0D4,10821
@@ -21,27 +21,27 @@ tests/kernels/ragged_paged_attention_kernel_v3_test.py,sha256=Hrd8iUkS1pS3rxeTyY
 tests/lora/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tests/lora/conftest.py,sha256=EXjwE1CjmUUlMEXpyE3UwxvgrKUllE73I8BNKfP1FTc,984
 tests/lora/test_bgmv.py,sha256=gQxWsJdNX2nkrE2xyrG0exwf3E2eHm2k2nkEXoANuQc,1359
-tests/lora/test_layers.py,sha256=21ekYlsK36r1GPZOfzs7E-KIsfI1JcuZl1E6vaQbHf4,26273
+tests/lora/test_layers.py,sha256=6B4HhMAItQmt0hPAQgyXgwSYs7b3bIbUf6LaPsqXLzY,25923
 tests/lora/test_lora.py,sha256=wJiF1P1BDnPN8TLX2tlFtdZ_QCkV-S9nPl6_uR6DqFc,4439
-tests/lora/utils.py,sha256=dR_v1H20vPVjFHdBhDajWOz0WJZlKuPLgMFQsME0LtA,3009
-tpu_inference/__init__.py,sha256=7IduGWw-_fwx0VA6EvC_AqHF67fnnShz6YvkqCfvFx8,1317
+tests/lora/utils.py,sha256=rY0tDZEZe58ye4-ykwrTnsiWuLcaEG57N_Rua90bDXI,2726
+tpu_inference/__init__.py,sha256=p4MaepRdN7723FUNE-3pOMxZWjFn4_TVFgjrNyty4JE,2304
 tpu_inference/env_override.py,sha256=pmL7lfs_rGCP92ya3wuWuudsCYeOMZ6tFZY82A4KkQc,365
-tpu_inference/envs.py,sha256=MTT_Pdtd6cAcciYjv1OekEmvspaq3SYL0oR_jDkQ_aE,3948
+tpu_inference/envs.py,sha256=ugze6VdQ_hG1IxUCbcgXZq7a22fZ-Lora3V_fkFOefw,5714
 tpu_inference/logger.py,sha256=HQCz7NefmbturuhOC7-3Ixbtcdgoz4g9FHh2RB6o8cc,334
-tpu_inference/tpu_info.py,sha256=9UohshkndR6dZpGWpWXfTD4qvIVdVgHf0yOoSEkLTrw,2276
-tpu_inference/utils.py,sha256=iGPY147jP_8AKMu3g7vYTndjJJiOrK_4opA0JWtws5Q,10068
+tpu_inference/tpu_info.py,sha256=3iilHRQSFjwMJwhKcuuawTm7mhwkgHbj4zi6CiAySrs,2265
+tpu_inference/utils.py,sha256=mHbjI8fxInPxagLsSUg-R3DzSz-X7WYNdoorPYoE3hg,10855
 tpu_inference/core/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/core/core_tpu.py,sha256=WDD3koE_j1QhWS2BbMA2aQOZayPZm4tYPvzL4YCX2jY,33294
 tpu_inference/core/disagg_executor.py,sha256=HZpgYMVxRxm0RQxO4l8IDYBWJ6Z3Tac6xavc5otcirc,4657
-tpu_inference/core/disagg_utils.py,sha256=ufWNFWQ5n4YnZpPOtoReHlYo4dlN7AbIqCyqS4an0t4,1572
+tpu_inference/core/disagg_utils.py,sha256=lv8MAVoAjtcmTaenUXVokg2q3d0tzsma86UiQlQ3omY,1492
 tpu_inference/core/sched/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/core/sched/dp_scheduler.py,sha256=mKs8Ms46szdlBfo8hjdqis2ZKAZbcKnHAGfEr0X5R8g,22527
 tpu_inference/distributed/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/distributed/jax_parallel_state.py,sha256=5_xCwcL03lFPUoSO_OP7hIVKpUFroW1m-jVO7R6FbUc,2223
-tpu_inference/distributed/tpu_connector.py,sha256=Zah46Sm5iOuh72SzXw69NxMc0MLnqsLEpe2BfDhpnqA,29731
-tpu_inference/distributed/utils.py,sha256=RwFQi8G4TzN1g9RjQu0pb5JxSc_jhoIZVsFJo0uHjxo,1513
+tpu_inference/distributed/tpu_connector.py,sha256=kLaTwy6BrAThJeFkd1soJ47bBo5iGp4GjUJs7xFx4Tg,29696
+tpu_inference/distributed/utils.py,sha256=1KIREn28Zg10O-MSUkVQMRzS09WoGc_VLGOX4QTFJac,1504
 tpu_inference/executors/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tpu_inference/executors/ray_distributed_executor.py,sha256=UgJP-XSgDPKDj_mkVQ16XrRN96juVpnFl6fdWEyFL_Y,15249
+tpu_inference/executors/ray_distributed_executor.py,sha256=9CnzWb8aurH1B0tJfMHB73F-RQBGqSf5DnymetBvZ5o,16225
 tpu_inference/experimental/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/experimental/llama3_jax_stashed.py,sha256=YK1oSIfto9ALo-HB45XfSrbq9XgVbE4m2C-9zRwmSzI,10913
 tpu_inference/kernels/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -53,7 +53,7 @@ tpu_inference/kernels/flash_attention/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCe
 tpu_inference/kernels/flash_attention/kernel.py,sha256=n8gmAFVfchMXlyaSEj8xXJm6AadFt26edQihPRdithY,25897
 tpu_inference/kernels/fused_moe/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/kernels/fused_moe/v1/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tpu_inference/kernels/fused_moe/v1/kernel.py,sha256=QHB0QEvC3x_6zhwz06JQpaOncQcNAhOSV92dD5tGVq8,40869
+tpu_inference/kernels/fused_moe/v1/kernel.py,sha256=xVXfclgbw_3U7c5W1azDFkFDK5FolBzDN9IL0rIzLQs,62813
 tpu_inference/kernels/mla/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/kernels/mla/v1/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/kernels/mla/v1/kernel.py,sha256=dw1nhpL47uQxMFOIN2kENC6aITbalT81YZLAyr1usLU,51571
@@ -67,18 +67,18 @@ tpu_inference/kernels/ragged_paged_attention/v2/kernel.py,sha256=OiQGAHhyggbp1Pe
 tpu_inference/kernels/ragged_paged_attention/v2/ragged_kv_cache_update.py,sha256=vGp2ZWODTbjyG9z2z0Qf_BX-wYHd5bUybnc_DtOz0nI,10995
 tpu_inference/kernels/ragged_paged_attention/v2/tuned_block_sizes.py,sha256=mw80bXBGenroGdrITV0F_EaI2s-Z9KWwqU9WodvJg14,97919
 tpu_inference/kernels/ragged_paged_attention/v3/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tpu_inference/kernels/ragged_paged_attention/v3/kernel.py,sha256=tlP6121yfXaukx_RQroHlHcZnbKPyyum0lAcvT0B_Pk,56132
-tpu_inference/kernels/ragged_paged_attention/v3/kernel_hd64.py,sha256=pD1Pte3neoLAxE3I3-VyV_4FuqgCHeAHGzEjMVt0MMk,56004
+tpu_inference/kernels/ragged_paged_attention/v3/kernel.py,sha256=O179Fft5KpuN5LIFx3SghWXJJUqh3Og-xqfO4Z8QXYU,57032
+tpu_inference/kernels/ragged_paged_attention/v3/kernel_hd64.py,sha256=ArwrqIQiKIop_jaDKAMw656YHQ3IFZ0sRu9Cgycrtko,59858
 tpu_inference/kernels/ragged_paged_attention/v3/tuned_block_sizes.py,sha256=k3LwduhZO85cJ-pSgnGN0c2Nn8eNeQq4eA94KUXJzMw,142198
 tpu_inference/kernels/ragged_paged_attention/v3/tuned_block_sizes_hd64.py,sha256=P3_ivi8iUz5QMU_3pgpl4Bkbmn0q0NpDtVJX39haRQA,11208
 tpu_inference/kernels/ragged_paged_attention/v3/util.py,sha256=1N_ozjKboDYLteFJndWoLXNudj2z53rGXMkELa5Z9tY,1102
 tpu_inference/layers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/layers/common/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tpu_inference/layers/common/attention_interface.py,sha256=CImMS8tuWgvaRY9YbGS3pY7OBnzeJ4Jla7LRFb4Xoa4,13224
+tpu_inference/layers/common/attention_interface.py,sha256=SQZ-1I32Jqg7GGI-z4BVibXbaitJHyTs26X3B5nBRVo,13369
 tpu_inference/layers/common/attention_metadata.py,sha256=St8ZatbY1D7xQACKJH459jMgp3oTP3AQ36mi9FZdrPU,850
 tpu_inference/layers/common/binary_search.py,sha256=ZQi-z1wG6WTcfVQXeTGOZokX4K1DSf9kCzqfrhEU8lk,12320
 tpu_inference/layers/common/quant_methods.py,sha256=mQSxZ44-QQtm22C_8ViejnP1cP2Dv6yc2YaP6oMKJeQ,185
-tpu_inference/layers/common/sharding.py,sha256=wBqdkXZSWfnnH8pkJtyW2DSqmAe_V4Vxi0iMPaXq0Z0,25185
+tpu_inference/layers/common/sharding.py,sha256=sjbwkDr2fP26Ob8f5cSDeDifr3eWFZMDHU4MKr7pIgQ,25217
 tpu_inference/layers/jax/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/layers/jax/base.py,sha256=Vhts6ZMwNCZ8LbnEXeB0rl3nHdS5hDJWX7HEa7Fl7yE,5775
 tpu_inference/layers/jax/constants.py,sha256=NcYg0zAf3ClfP7YMYdYu_F1GngOzZaIxIAHBZDunKw4,2755
@@ -102,14 +102,14 @@ tpu_inference/layers/jax/sample/sampling.py,sha256=C30KgmdOVSaagvHhbfLgVJtVQmJo8
 tpu_inference/layers/jax/sample/sampling_metadata.py,sha256=Gd835LNWfGM0NRQBVBqEv0nPwt5q9F4AdFym0CUS1fw,2561
 tpu_inference/layers/vllm/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/layers/vllm/attention.py,sha256=wbJpcgqEAuIirv5PIULbiP-ggMKjmTanbB7Dg0BVYv4,7366
-tpu_inference/layers/vllm/fused_moe.py,sha256=XZt2CPUz00qZzDcyfBFz6buhVzmGL1amHalHJALl9zw,18945
+tpu_inference/layers/vllm/fused_moe.py,sha256=5rzPzMsKRmsD-hh2Rue0ahaeyhh6cHB7FUIrKfn4s0U,21089
 tpu_inference/layers/vllm/linear_common.py,sha256=_YlJtbdaYcck_j-gFLos_k0ycktVWxT8Qo57tR2YqJ8,7749
-tpu_inference/layers/vllm/sharding.py,sha256=WTx1tF_7R99AdyE-lL7HQJ378hAafeI-JVRsugAvwn4,9177
+tpu_inference/layers/vllm/sharding.py,sha256=as7CF8UKTF3ToymwRY5Pi8uzwJk0P1sHPkWB5xEx3mA,9169
 tpu_inference/layers/vllm/quantization/__init__.py,sha256=SEppGayBzzQ5tsXLSy99aqilkAawQwYxnv2alCg6-ZU,1777
 tpu_inference/layers/vllm/quantization/awq.py,sha256=-8ZmjGvSKJB6_JuwSctNWt8xHWq4VSvK_AK9iahlgCo,8495
-tpu_inference/layers/vllm/quantization/common.py,sha256=wm3pge6XMTMsLK7_SSdgBP0PvQzz-1mrqN2I6xMqzrc,4218
-tpu_inference/layers/vllm/quantization/mxfp4.py,sha256=KwGoqIiPkd6FplGuYAKi4uX5A8MPlZqq99MVPchXyi4,11561
-tpu_inference/layers/vllm/quantization/unquantized.py,sha256=q9w_q4KmKw5TltzU48xWjEEslYoIjkSbxq_wmAxaLjw,15920
+tpu_inference/layers/vllm/quantization/common.py,sha256=8XD64pPa077c9HThFhLFVHlDL9YBafnYwp6rp6gR44E,4432
+tpu_inference/layers/vllm/quantization/mxfp4.py,sha256=o661uiSvLvWGr8hQMl7TqYXJyALPREtNWlKHAM9AUrw,14541
+tpu_inference/layers/vllm/quantization/unquantized.py,sha256=nSRBzVurTiQQkF9FuSTshfRwfxfzs54E2_4eK7Eyhj0,15345
 tpu_inference/layers/vllm/quantization/compressed_tensors/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/layers/vllm/quantization/compressed_tensors/compressed_tensors.py,sha256=6idEyy3e849fZ1UeNvc9eSHYX7e6qvohrJa_d_D9MBk,5285
 tpu_inference/layers/vllm/quantization/compressed_tensors/compressed_tensors_moe.py,sha256=FM901QhyhJRC8CuMeICzCVVERvBHbhruRxYW0EQ570s,8820
@@ -118,62 +118,57 @@ tpu_inference/layers/vllm/quantization/compressed_tensors/schemes/compressed_ten
 tpu_inference/layers/vllm/quantization/compressed_tensors/schemes/compressed_tensors_w8a8_int8.py,sha256=4y7lYgybpXszpCAtxGFhR8LDEbEoCCeo3DfUSOXxhaQ,5202
 tpu_inference/lora/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/lora/torch_lora_ops.py,sha256=pr3N7DVfkn3ANijUC6dBoiCtIJW4fdJpKdC3zWBUsxE,3121
-tpu_inference/lora/torch_punica_tpu.py,sha256=b27DpmIS_N5bhlIcryiENYNmPxp_cu40CGxjPW64d44,12706
-tpu_inference/mock/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tpu_inference/mock/vllm_config_utils.py,sha256=FlQshLjoHdgs3C66tYHYbKFUjbk9DhUwY-7HibZk0fI,878
-tpu_inference/mock/vllm_envs.py,sha256=cCubeOhH2WeYZQFJt6W0y_IiQo0fzIWR1LCCE8i6kI4,50990
-tpu_inference/mock/vllm_logger.py,sha256=vUGnN5nKT--ZvU15YCzODUM_FGiXKhcrrjDGjeN00RQ,7297
-tpu_inference/mock/vllm_logging_utils.py,sha256=TEUmKj3xHiLzHBnFqAujcxH0t2hBQ04sUaho2RyORnk,486
+tpu_inference/lora/torch_punica_tpu.py,sha256=qTnXZGLoOgvukSxeunO_SfpPTlkq9GlMj9H7zVYg9LE,12680
 tpu_inference/models/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/models/common/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tpu_inference/models/common/model_loader.py,sha256=YF7nlglitwy2u8bexnzz_urTlsjIije-QWaWbaka5O8,18228
+tpu_inference/models/common/model_loader.py,sha256=b3aigca81gMVJt42oF2aoRohQHjBBe3oK3IPblZAaUM,19996
 tpu_inference/models/jax/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/models/jax/deepseek_v3.py,sha256=SKOHVEC-_2NLxBnzBzbu5tu0d6FTlAEiI1EefGaO2QE,40047
 tpu_inference/models/jax/gpt_oss.py,sha256=Vw4LRB5Kp6hbA2hjZGFS8kiEqOCjf881XH2JNtu2S1I,20924
 tpu_inference/models/jax/jax_intermediate_tensor.py,sha256=Pxu1PCV5LN5X58aYVkPiohcXZIeKVim2oqvrS_cVgw4,2604
-tpu_inference/models/jax/llama3.py,sha256=w99DAfipGS9HyX2ZRwqyYLxC3oa0ew5eEQ6EXlMMf18,13426
+tpu_inference/models/jax/llama3.py,sha256=ZiFtrpAzXTT9vAPES9UeuJInCWGbvDWs7g0_JLdCCa4,13479
 tpu_inference/models/jax/llama4.py,sha256=wf2Sp2iYViaYD5rSfv3_ryO6gYuYM5XaOyvghaP4OCY,29631
-tpu_inference/models/jax/llama_eagle3.py,sha256=STUkAK6XEA7JM3i_Lx36-t5BhkAGeW_xYiq3zYhHP1A,12297
-tpu_inference/models/jax/phi3.py,sha256=TpP3Nvr1myW_Qd8xNrLP1VmXtq7BuTcWNayJitskFd0,13579
-tpu_inference/models/jax/qwen2.py,sha256=P_x_Qygf-nanmF8Uufk4c-qLNxP4RAk4yuqSF8VwbxE,13357
-tpu_inference/models/jax/qwen2_5_vl.py,sha256=fvMgM5GfUn5EECaMbR0z37mmbCHphAT1AvWPvGkhVn4,43942
-tpu_inference/models/jax/qwen3.py,sha256=lr3TIIQKmNgWFDFxwuPsVOypqBijkqrpnNCopVg4iBo,10997
+tpu_inference/models/jax/llama_eagle3.py,sha256=7-U99yvBkle-FSZ3NDDI-obWSQ2Fo2OTOi1H67H4jxY,12476
+tpu_inference/models/jax/llama_guard_4.py,sha256=LrnU2zBWM0s4q_5dwmR--OO0V7ttltsYhrHYlBgQVIw,15275
+tpu_inference/models/jax/qwen2.py,sha256=SuAp7tErk8OoIRko0Vt6QSOZP_9B9r5GTfqmVfImUIo,13410
+tpu_inference/models/jax/qwen2_5_vl.py,sha256=WUOmqNE6fHQ8PGU85Y8Bt6-CtCC1Uubbox_9FdpDMMo,49833
+tpu_inference/models/jax/qwen3.py,sha256=CIZQKjZDke_LPGsLNhRCJdDTzWueUneBPAQ1blS24IM,11050
 tpu_inference/models/jax/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/models/jax/utils/file_utils.py,sha256=NOuSC3YFnZpf3CZgYdghbbiNYJt42zgjlEYbOZIVct4,2840
 tpu_inference/models/jax/utils/multi_modal_utils.py,sha256=rrIrQWidkUnGilBHKNpdYh7_2BkvnAaqanXjC81GNcg,6156
-tpu_inference/models/jax/utils/weight_utils.py,sha256=65-H8BTbyilIBMBfvWjkkW3mf4soYASbhrJFqbFKzL4,20129
+tpu_inference/models/jax/utils/weight_utils.py,sha256=qFU53jPHPvIcs_EOdIH80oNojpUp7GdSY2E6NZNsjvM,21376
 tpu_inference/models/jax/utils/quantization/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/models/jax/utils/quantization/mxfp4_utils.py,sha256=boGnqJCRIOf5nedAxQ8_IUTV6Rfll10DXnRC40BeeE8,3682
-tpu_inference/models/jax/utils/quantization/quantization_utils.py,sha256=xgKoKB7AM3TYPxzVgEGLTK9ebQH2Kx8mNuO0heovkmk,26778
+tpu_inference/models/jax/utils/quantization/quantization_utils.py,sha256=rzAFU3OtQvg8w8ow0V15rMljAsa4SBrwOye6OI8Bty4,26530
 tpu_inference/models/jax/utils/quantization/configs/fp8_all_modules_w_only.yaml,sha256=d_YHPtaRJ_7PBrPijSzJGnVeoJO62tKIGqrgFqpYT1k,137
 tpu_inference/models/jax/utils/quantization/configs/fp8_default.yaml,sha256=b7SyL75HuSTj3fN9_ZLCK_CDiccL5DGq_DddGmxj_qk,170
 tpu_inference/models/jax/utils/quantization/configs/int8_all_modules_w_only.yaml,sha256=0Qwij71zj9k6rmrUNd8Q5df9YYfkoJ1ZkgMAHxQy81k,128
 tpu_inference/models/jax/utils/quantization/configs/int8_default.yaml,sha256=lGec0UwwxmNPNgKPSsTsCMSXNJjhw507KMtM2NsSCMw,152
 tpu_inference/models/vllm/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tpu_inference/models/vllm/vllm_model_wrapper.py,sha256=BREx7psxRid0z8tyLflF12ZBCxXgZqnBR9SAdGm8aVg,12001
+tpu_inference/models/vllm/vllm_model_wrapper.py,sha256=3EcaD_1vZuyAZBfDtm5u_qfCahQU28qR4rAUraNAFqs,12305
 tpu_inference/models/vllm/vllm_model_wrapper_context.py,sha256=yxlJHPmRQIAwlb1MmHK3xfXokgIkJ-evNU4PgyoJUdg,1187
 tpu_inference/platforms/__init__.py,sha256=lQCrKddS_GcGpCbeogvz9zOZD1mQw5bBsiw8On46qFQ,74
-tpu_inference/platforms/tpu_platform.py,sha256=AYFr1Q7VUN76wcdgOe_wZuVIHgp2U8isBJ3iHrYqt0M,10530
+tpu_inference/platforms/tpu_platform.py,sha256=F4jjPEFHFUTxdfWZYTBuUVJt6SYTFeWEKmrl74sX-Zk,10663
 tpu_inference/runner/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/runner/block_table.py,sha256=K3Ic8EgPM08d_C5nEN60mxoRydlaQWySAemf_8Q_qVw,4175
-tpu_inference/runner/compilation_manager.py,sha256=WvpbHzSAW7_rNJE6M8YIr6BLYoeZraBzncl1Pr0RfdY,36180
+tpu_inference/runner/compilation_manager.py,sha256=dU0Yk8f0LtRTBe2q0iB3xcMSRco_WPsj2wS6zZJ8WhY,40375
 tpu_inference/runner/input_batch.py,sha256=bx221NX2IOWzrtopss-B-2ZKW4y-U6nQpG09PjpUziw,18273
 tpu_inference/runner/kv_cache.py,sha256=F4dzW2d53xuxkFUn0oKzwE6VklGUeVm-QM19NVfIQDU,4577
-tpu_inference/runner/kv_cache_manager.py,sha256=CJxXtdWuewJqcTBMoR70_Uvwxjtc3cK2jxe1KpI9kQc,22152
+tpu_inference/runner/kv_cache_manager.py,sha256=N0a896CE7Zrs_d4ZSSzRdqgjV1It57RBDSIpOzkRqro,22013
 tpu_inference/runner/lora_utils.py,sha256=B4xMCgXGJ4VNdePvn89HH3tIZ-gYsQ7Vq_YCiYIATEY,3843
 tpu_inference/runner/multimodal_manager.py,sha256=azEPdHOwz8CN11MQmorGdtrCLbFaTCxdWyuEsZTzjYM,9778
-tpu_inference/runner/persistent_batch_manager.py,sha256=KERSfKy6XjMejnbtPGI3hzoYAHJLeCxmpZVYPqBCago,11156
+tpu_inference/runner/persistent_batch_manager.py,sha256=Otu67vOTf1_HKAMZgPDDHlRvvZ3YVJdz-QderH4qOII,13263
 tpu_inference/runner/speculative_decoding_manager.py,sha256=I3FDWKh2dn6nV8LgTGfCTwMKYnxQsTPpBIrmaJngXHs,10215
-tpu_inference/runner/structured_decoding_manager.py,sha256=Y0ERPhj4olFh6Y2TxP0R1_4UIJwy7nemYA-h63YIR2U,3622
-tpu_inference/runner/tpu_runner.py,sha256=3SZYn0CBA4LOaTO3GdQOxKx3HKmVcNmUEeSyzSAGyFY,73320
-tpu_inference/runner/utils.py,sha256=ZnWUoNo-7INeB0mdXti1jwUOdbmxyExznOs-crRTQLk,17126
+tpu_inference/runner/structured_decoding_manager.py,sha256=gZQKQUFxh6xYYH9eGTdbguqk8hc2WwTrIdMMuCcbymE,3573
+tpu_inference/runner/tpu_runner.py,sha256=NBDKfSGShHmYpudrtGfo1hnVSQTcLpZV_nPiXEo7JPQ,79439
+tpu_inference/runner/utils.py,sha256=lKqL5nxGTk7ufzJRNdp4udn2bPu3jIX52W7akXgSrHc,17133
 tpu_inference/spec_decode/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tpu_inference/spec_decode/jax/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tpu_inference/spec_decode/jax/eagle3.py,sha256=A1dt-dmBttpy-5DGcL4noEDCB0OGP8Xo6MXqgJvWIo8,16593
+tpu_inference/spec_decode/jax/eagle3.py,sha256=FxP0uWeQlHlgCpt1nY3FUd4lKlegKJljHyc05jJucaQ,19104
 tpu_inference/worker/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-tpu_inference/worker/tpu_worker.py,sha256=0ZguK2BtIQjQSvyUTcUH9ENBrxt09w3CbgPoDY13Eok,14210
-tpu_inference-0.11.1.dev202511150811.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-tpu_inference-0.11.1.dev202511150811.dist-info/METADATA,sha256=jwcAlddP8qvhnZDlYn5YqhKiJW2TetHR5ITY6M3zSic,5465
-tpu_inference-0.11.1.dev202511150811.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-tpu_inference-0.11.1.dev202511150811.dist-info/top_level.txt,sha256=gb1hRIQ3DOawUfVzvPL2E__2KPIl9I0vb5r0xcRBGYQ,20
-tpu_inference-0.11.1.dev202511150811.dist-info/RECORD,,
+tpu_inference/worker/tpu_worker.py,sha256=LnZcSNxdhh0NkoWXxS5bZ0bsTMduSANehy2wELAaVsY,20672
+tpu_inference-0.11.1.dev202512030818.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+tpu_inference-0.11.1.dev202512030818.dist-info/METADATA,sha256=oLzYFTCTvHDQLfyWoc8qV4IMYCoLRTiHECf08oT_bFA,5517
+tpu_inference-0.11.1.dev202512030818.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+tpu_inference-0.11.1.dev202512030818.dist-info/top_level.txt,sha256=gb1hRIQ3DOawUfVzvPL2E__2KPIl9I0vb5r0xcRBGYQ,20
+tpu_inference-0.11.1.dev202512030818.dist-info/RECORD,,

tpu_inference/mock/vllm_config_utils.py

@@ -1,28 +0,0 @@
-from dataclasses import dataclass, field
-from typing import Any, List, Mapping
-
-
-@dataclass
-class ModelConfig():
-    max_model_len: int = 2048
-    max_prefill_len: int = 1024
-    prefill_batch_size: int = 1
-    decode_batch_size: int = 1
-    block_size: int = 16
-    num_layers: int = 32
-    num_kv_heads: int = 32
-    head_dim: int = 128
-    vocab_size: int = 32000
-    model: str = "llama3"
-    hf_config: str = ""
-    architectures: List[str] = field(default_factory=list)
-    override_generation_config: dict[str, Any] = field(default_factory=dict)
-    hf_overrides: dict[str, Any] = field(default_factory=dict)
-
-
-@dataclass
-class VllmConfig():
-    additional_config: Mapping[str, Any] = field(default_factory=dict)
-    # Set default max_model_len to turn off warnings.
-    model_config: ModelConfig = field(
-        default_factory=lambda: ModelConfig(max_model_len=1024))