ipex-llm 2.3.0b20250413__py3-none-win_amd64.whl → 2.3.0b20250415__py3-none-win_amd64.whl

ipex_llm/transformers/convert.py

@@ -1089,6 +1089,15 @@ def ggml_convert_low_bit(model, qtype, optimize_model=True,
                          embedding_qtype=None,
                          mixed_precision=False,
                          disable_optimize_pre=False):
+    if qtype == ggml_tensor_qtype["sym_int4"] and torch.__version__ >= "2.6":
+        logger.warning("sym_int4 is deprecated, use woq_int4 instead, "
+                       "if you are loading saved sym_int4 low bit model, "
+                       "please resaved it with woq_int4")
+        qtype = ggml_tensor_qtype["woq_int4"]
+    elif qtype == ggml_tensor_qtype["woq_int4"] and torch.__version__ < "2.6":
+        logger.warning("woq_int4 is not supported with pytorch<2.6, "
+                       "use sym_int4 instead or use ipex-llm with pytorch>=2.6")
+        qtype = ggml_tensor_qtype["sym_int4"]
     if qtype in ggml_tensor_qtype.values():
         index = list(ggml_tensor_qtype.values()).index(qtype)
         logger.info(f"Converting the current model to "

ipex_llm/transformers/models/baichuan.py

@@ -326,14 +326,17 @@ def baichuan_attention_forward_13b(
         else:
             attention_mask = attention_mask[None, :, -q_len:, :]
 
+    head_dim = query_states.shape[-1]
+    scale = 1 / math.sqrt(head_dim)
+
     if use_sdp(q_len, kv_seq_len, self.head_dim, query_states):
         import xe_addons
         if use_quantize_kv:
             attn_output = xe_addons.sdp_fp8(query_states, key_states, value_states,
-                                            attention_mask)
+                                            attention_mask, scale)
         else:
             attn_output = xe_addons.sdp(query_states, key_states, value_states,
-                                        attention_mask)
+                                        attention_mask, scale)
         attn_weights = None
     else:
         if use_quantize_kv:

ipex_llm/transformers/models/chatglm.py

@@ -68,7 +68,9 @@ def glm_sdpa(query, key, value, attention_mask=None, is_causal=False):
         if use_sdp(query.shape[2], key.shape[2],
                    query.shape[-1], query):
             import xe_addons
-            attn_output = xe_addons.sdp(query, key, value, attn_bias)
+            head_dim = query.shape[-1]
+            scale = 1 / math.sqrt(head_dim)
+            attn_output = xe_addons.sdp(query, key, value, attn_bias, scale)
             context_layer = attn_output.view(query.shape)
         else:
             head_dim = query.size(-1)

ipex_llm/transformers/models/qwen_vl.py

@@ -164,7 +164,9 @@ def qwen_attention_forward_vl(
     if not self.training and not hidden_states.requires_grad and \
             use_sdp(q_len, key.shape[2], self.head_dim, query):
         import xe_addons
-        attn_output = xe_addons.sdp(query, key, value, attention_mask)
+        head_dim = query.shape[-1]
+        scale = 1 / math.sqrt(head_dim)
+        attn_output = xe_addons.sdp(query, key, value, attention_mask, scale)
         attn_output = attn_output.view(query.shape)
         attn_output = attn_output.transpose(1, 2)
         attn_weight = None

{ipex_llm-2.3.0b20250413.dist-info → ipex_llm-2.3.0b20250415.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: ipex-llm
-Version: 2.3.0b20250413
+Version: 2.3.0b20250415
 Summary: Large Language Model Develop Toolkit
 Home-page: https://github.com/intel-analytics/ipex-llm
 Author: BigDL Authors
@@ -27,7 +27,7 @@ Requires-Dist: intel-openmp ; (platform_machine == "x86_64" or platform_machine
 Requires-Dist: torch ==2.1.2+cpu ; (platform_system == "Linux") and extra == 'all'
 Requires-Dist: torch ==2.1.2 ; (platform_system == "Windows") and extra == 'all'
 Provides-Extra: cpp
-Requires-Dist: bigdl-core-cpp ==2.7.0b20250413 ; extra == 'cpp'
+Requires-Dist: bigdl-core-cpp ==2.7.0b20250415 ; extra == 'cpp'
 Requires-Dist: setuptools ; extra == 'cpp'
 Requires-Dist: onednn-devel ==2025.0.1 ; (platform_system == "Windows") and extra == 'cpp'
 Requires-Dist: onednn ==2025.0.1 ; (platform_system == "Windows") and extra == 'cpp'
@@ -60,7 +60,7 @@ Requires-Dist: transformers ==4.40.0 ; extra == 'npu'
 Requires-Dist: intel-openmp ; (platform_machine == "x86_64" or platform_machine == "AMD64") and extra == 'npu'
 Requires-Dist: torch ==2.1.2+cpu ; (platform_system == "Linux") and extra == 'npu'
 Requires-Dist: torch ==2.1.2 ; (platform_system == "Windows") and extra == 'npu'
-Requires-Dist: bigdl-core-npu ==2.7.0b20250413 ; (platform_system == "Windows") and extra == 'npu'
+Requires-Dist: bigdl-core-npu ==2.7.0b20250415 ; (platform_system == "Windows") and extra == 'npu'
 Provides-Extra: serving
 Requires-Dist: py-cpuinfo ; extra == 'serving'
 Requires-Dist: fschat[model_worker,webui] ==0.2.36 ; extra == 'serving'
@@ -80,9 +80,9 @@ Requires-Dist: setuptools <70.0.0 ; extra == 'xpu'
 Requires-Dist: torch ==2.1.0a0 ; extra == 'xpu'
 Requires-Dist: torchvision ==0.16.0a0 ; extra == 'xpu'
 Requires-Dist: intel-extension-for-pytorch ==2.1.10+xpu ; extra == 'xpu'
-Requires-Dist: bigdl-core-xe-21 ==2.7.0b20250413 ; extra == 'xpu'
-Requires-Dist: bigdl-core-xe-batch-21 ==2.7.0b20250413 ; extra == 'xpu'
-Requires-Dist: bigdl-core-xe-addons-21 ==2.7.0b20250413 ; extra == 'xpu'
+Requires-Dist: bigdl-core-xe-21 ==2.7.0b20250415 ; extra == 'xpu'
+Requires-Dist: bigdl-core-xe-batch-21 ==2.7.0b20250415 ; extra == 'xpu'
+Requires-Dist: bigdl-core-xe-addons-21 ==2.7.0b20250415 ; extra == 'xpu'
 Provides-Extra: xpu-2-1
 Requires-Dist: py-cpuinfo ; extra == 'xpu-2-1'
 Requires-Dist: protobuf ; extra == 'xpu-2-1'
@@ -97,9 +97,9 @@ Requires-Dist: setuptools <70.0.0 ; extra == 'xpu-2-1'
 Requires-Dist: torch ==2.1.0a0 ; extra == 'xpu-2-1'
 Requires-Dist: torchvision ==0.16.0a0 ; extra == 'xpu-2-1'
 Requires-Dist: intel-extension-for-pytorch ==2.1.10+xpu ; extra == 'xpu-2-1'
-Requires-Dist: bigdl-core-xe-21 ==2.7.0b20250413 ; extra == 'xpu-2-1'
-Requires-Dist: bigdl-core-xe-batch-21 ==2.7.0b20250413 ; extra == 'xpu-2-1'
-Requires-Dist: bigdl-core-xe-addons-21 ==2.7.0b20250413 ; extra == 'xpu-2-1'
+Requires-Dist: bigdl-core-xe-21 ==2.7.0b20250415 ; extra == 'xpu-2-1'
+Requires-Dist: bigdl-core-xe-batch-21 ==2.7.0b20250415 ; extra == 'xpu-2-1'
+Requires-Dist: bigdl-core-xe-addons-21 ==2.7.0b20250415 ; extra == 'xpu-2-1'
 Requires-Dist: intel-openmp ; (platform_machine == "x86_64" or platform_machine == "AMD64") and extra == 'xpu-2-1'
 Requires-Dist: dpcpp-cpp-rt ==2024.0.2 ; (platform_system == "Windows") and extra == 'xpu-2-1'
 Requires-Dist: mkl-dpcpp ==2024.0.0 ; (platform_system == "Windows") and extra == 'xpu-2-1'
@@ -117,7 +117,7 @@ Requires-Dist: setuptools ; extra == 'xpu-2-6'
 Requires-Dist: torch ==2.6.0+xpu ; extra == 'xpu-2-6'
 Requires-Dist: torchvision ==0.21.0+xpu ; extra == 'xpu-2-6'
 Requires-Dist: torchaudio ==2.6.0+xpu ; extra == 'xpu-2-6'
-Requires-Dist: bigdl-core-xe-all ==2.7.0b20250413 ; extra == 'xpu-2-6'
+Requires-Dist: bigdl-core-xe-all ==2.7.0b20250415 ; extra == 'xpu-2-6'
 Requires-Dist: onednn-devel ==2025.0.1 ; extra == 'xpu-2-6'
 Requires-Dist: onednn ==2025.0.1 ; extra == 'xpu-2-6'
 Requires-Dist: dpcpp-cpp-rt ==2025.0.2 ; extra == 'xpu-2-6'
@@ -132,7 +132,7 @@ Requires-Dist: tokenizers ==0.15.2 ; extra == 'xpu-2-6-arl'
 Requires-Dist: accelerate ==0.23.0 ; extra == 'xpu-2-6-arl'
 Requires-Dist: tabulate ; extra == 'xpu-2-6-arl'
 Requires-Dist: setuptools ; extra == 'xpu-2-6-arl'
-Requires-Dist: bigdl-core-xe-all ==2.7.0b20250413 ; extra == 'xpu-2-6-arl'
+Requires-Dist: bigdl-core-xe-all ==2.7.0b20250415 ; extra == 'xpu-2-6-arl'
 Requires-Dist: onednn-devel ==2025.0.1 ; extra == 'xpu-2-6-arl'
 Requires-Dist: onednn ==2025.0.1 ; extra == 'xpu-2-6-arl'
 Requires-Dist: dpcpp-cpp-rt ==2025.0.2 ; extra == 'xpu-2-6-arl'
@@ -155,9 +155,9 @@ Requires-Dist: tokenizers ==0.15.2 ; extra == 'xpu-arc'
 Requires-Dist: accelerate ==0.23.0 ; extra == 'xpu-arc'
 Requires-Dist: tabulate ; extra == 'xpu-arc'
 Requires-Dist: setuptools ; extra == 'xpu-arc'
-Requires-Dist: bigdl-core-xe-23 ==2.7.0b20250413 ; extra == 'xpu-arc'
-Requires-Dist: bigdl-core-xe-batch-23 ==2.7.0b20250413 ; extra == 'xpu-arc'
-Requires-Dist: bigdl-core-xe-addons-23 ==2.7.0b20250413 ; extra == 'xpu-arc'
+Requires-Dist: bigdl-core-xe-23 ==2.7.0b20250415 ; extra == 'xpu-arc'
+Requires-Dist: bigdl-core-xe-batch-23 ==2.7.0b20250415 ; extra == 'xpu-arc'
+Requires-Dist: bigdl-core-xe-addons-23 ==2.7.0b20250415 ; extra == 'xpu-arc'
 Requires-Dist: intel-openmp ; (platform_machine == "x86_64" or platform_machine == "AMD64") and extra == 'xpu-arc'
 Requires-Dist: torch ==2.3.1+cxx11.abi ; (platform_system == "Linux") and extra == 'xpu-arc'
 Requires-Dist: torchvision ==0.18.1+cxx11.abi ; (platform_system == "Linux") and extra == 'xpu-arc'
@@ -178,9 +178,9 @@ Requires-Dist: tokenizers ==0.15.2 ; extra == 'xpu-arl'
 Requires-Dist: accelerate ==0.23.0 ; extra == 'xpu-arl'
 Requires-Dist: tabulate ; extra == 'xpu-arl'
 Requires-Dist: setuptools ; extra == 'xpu-arl'
-Requires-Dist: bigdl-core-xe-23 ==2.7.0b20250413 ; extra == 'xpu-arl'
-Requires-Dist: bigdl-core-xe-batch-23 ==2.7.0b20250413 ; extra == 'xpu-arl'
-Requires-Dist: bigdl-core-xe-addons-23 ==2.7.0b20250413 ; extra == 'xpu-arl'
+Requires-Dist: bigdl-core-xe-23 ==2.7.0b20250415 ; extra == 'xpu-arl'
+Requires-Dist: bigdl-core-xe-batch-23 ==2.7.0b20250415 ; extra == 'xpu-arl'
+Requires-Dist: bigdl-core-xe-addons-23 ==2.7.0b20250415 ; extra == 'xpu-arl'
 Requires-Dist: intel-openmp ; (platform_machine == "x86_64" or platform_machine == "AMD64") and extra == 'xpu-arl'
 Requires-Dist: torch ==2.3.1+cxx11.abi ; (platform_system == "Linux") and extra == 'xpu-arl'
 Requires-Dist: torchvision ==0.18.1+cxx11.abi ; (platform_system == "Linux") and extra == 'xpu-arl'
@@ -201,9 +201,9 @@ Requires-Dist: tokenizers ==0.15.2 ; extra == 'xpu-lnl'
 Requires-Dist: accelerate ==0.23.0 ; extra == 'xpu-lnl'
 Requires-Dist: tabulate ; extra == 'xpu-lnl'
 Requires-Dist: setuptools ; extra == 'xpu-lnl'
-Requires-Dist: bigdl-core-xe-23 ==2.7.0b20250413 ; extra == 'xpu-lnl'
-Requires-Dist: bigdl-core-xe-batch-23 ==2.7.0b20250413 ; extra == 'xpu-lnl'
-Requires-Dist: bigdl-core-xe-addons-23 ==2.7.0b20250413 ; extra == 'xpu-lnl'
+Requires-Dist: bigdl-core-xe-23 ==2.7.0b20250415 ; extra == 'xpu-lnl'
+Requires-Dist: bigdl-core-xe-batch-23 ==2.7.0b20250415 ; extra == 'xpu-lnl'
+Requires-Dist: bigdl-core-xe-addons-23 ==2.7.0b20250415 ; extra == 'xpu-lnl'
 Requires-Dist: intel-openmp ; (platform_machine == "x86_64" or platform_machine == "AMD64") and extra == 'xpu-lnl'
 Requires-Dist: torch ==2.3.1+cxx11.abi ; (platform_system == "Linux") and extra == 'xpu-lnl'
 Requires-Dist: torchvision ==0.18.1+cxx11.abi ; (platform_system == "Linux") and extra == 'xpu-lnl'

{ipex_llm-2.3.0b20250413.dist-info → ipex_llm-2.3.0b20250415.dist-info}/RECORD

@@ -41,35 +41,35 @@ ipex_llm/langchain/llms/transformerspipelinellm.py,sha256=vm522YPPwWxxAPVvQBtxRf
 ipex_llm/langchain/vllm/__init__.py,sha256=T-EbRT6GJ_8RCu-iLmSzcftOimXSPQf2d5X72AUAy2Y,874
 ipex_llm/langchain/vllm/vllm.py,sha256=6dxc-ZISZQrJilEa_HA827l75Dv9rcHpY_G6FdJ8BVs,7793
 ipex_llm/libs/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-ipex_llm/libs/bloom-api.dll,sha256=dYVKVEjmanZsbHdMSq5TaQs45xZtNaOFBjhoK_9VFBo,36352
-ipex_llm/libs/bloom.dll,sha256=tuZSGxK-dHORSW6-XDbie-tFuFTtHz9lyMzstdRND1k,507904
-ipex_llm/libs/gptneox-api.dll,sha256=5QEyhSO0CUYc6Xl1RR4OAFwtVyf7KWU1Wx-22xPz3_U,24576
-ipex_llm/libs/gptneox.dll,sha256=Dn-56Nqwid5DcVnYbAJh-A4MgTw_D_WmL_iN1SBZGLI,568320
-ipex_llm/libs/libbloom_avx.dll,sha256=d0UjHHgrlSJgjw4VLyJSpjvzk1FxbpvJdwtFKbqujlw,536576
-ipex_llm/libs/libbloom_vnni.dll,sha256=nSwx260ErcCcDKfGy6PP8EOdfcJpvWUVsxFKHbS6wOY,508416
-ipex_llm/libs/libgptneox_avx.dll,sha256=sBf6S-EiIJcRLmhaKuuaadaU5fHQNVXVAB0VQJTLoRU,596992
-ipex_llm/libs/libgptneox_vnni.dll,sha256=LcNF7O5eyL2y7HMAU2LYJRY0w7NM59qJZ17qiuNsl24,568832
-ipex_llm/libs/libllama_avx.dll,sha256=bL1YY7uYu-JNklBk9cJF0gXzkU8Iy5PjN3NB5zU5xDY,591360
-ipex_llm/libs/libllama_vnni.dll,sha256=HOhU09jo6EjZmJcZSnt4UpAL_G1qKbRcAL5jpIESnaQ,563200
-ipex_llm/libs/libstarcoder_avx.dll,sha256=8w4KETb43ZPslHK2azxiGsv6pCsoXpNk-_bVi8lkTKc,627712
-ipex_llm/libs/libstarcoder_vnni.dll,sha256=yrOi0VdnHkJQUFFAq17_-Hy89PhL0NTjy4tiJtX0MP0,599552
-ipex_llm/libs/llama-api.dll,sha256=dMkSUtk-SB2nuZgZPtgmMa8CJ5IKgc8lH8F5RXJybW4,25600
-ipex_llm/libs/llama.dll,sha256=jfcQ_LJAMRHIwQsIc3ek1upKv5zGT0oJn5DlbjYw6ko,562688
-ipex_llm/libs/main-bloom.exe,sha256=xTMzHKGfg_85ytA-twHNwXbHIoyRmbj_SdGPCLDoSeg,103424
-ipex_llm/libs/main-gptneox.exe,sha256=aiqW57JxUfl4e4Q_qXbBu2EzlUBv2T2OVke3KUv57FY,98816
-ipex_llm/libs/main-llama.exe,sha256=4CBt_rQ6jgKtWXvkeL8WneIOTBFMhJNIm3uQjYim89Y,99840
-ipex_llm/libs/main-starcoder.exe,sha256=yn94YnDvvsfM_dhDuLw0RgMDr3XKJaZBIiuAc43hRyg,157696
-ipex_llm/libs/pipeline.dll,sha256=G_sGCdEwNB_D8V0AKK9MK7OrxhinqCcdyqEkzJWurwE,73216
-ipex_llm/libs/quantize-bloom.exe,sha256=D1VV8ytvXZqPXvdYNkctFBvHfP7tKH5uyzFePg1mp8M,126464
-ipex_llm/libs/quantize-bloom_vnni.exe,sha256=wfFVg69IwMMe42McXKlSmfbDOaSCYznKwZpQ8Hujhx0,128000
-ipex_llm/libs/quantize-gptneox.exe,sha256=WRtrYPGC1pI6Nj-nNQ07l8Z2YbH3Q0lLzjnYlZdWO2o,104448
-ipex_llm/libs/quantize-gptneox_vnni.exe,sha256=UDpeoMP2bWjG1ZAnEd95knk7BqUMlAbGDGE3oeODi14,104960
-ipex_llm/libs/quantize-llama.exe,sha256=a6yCAP_NrOGR2-GR1Wb_aaqWcamJT5MUpEm96vYGcB8,110080
-ipex_llm/libs/quantize-llama_vnni.exe,sha256=R1aRijj8oKx3KUVu3DTZ1vZGbYc3kcRD4SjYaY5XhcI,110592
-ipex_llm/libs/quantize-starcoder.exe,sha256=FDW9VGQ9-32noOGr4I81l5PETjBucwPC8r80uO5C8yw,127488
-ipex_llm/libs/quantize-starcoder_vnni.exe,sha256=fm7gka80MTxHd5HVroi_yTsFFU0yrcB5RMZL8BUqL8U,128512
-ipex_llm/libs/starcoder-api.dll,sha256=hfLMok4MNXXdD989vMNS83S6R5k-j3W1FG2Xrurg0SI,21504
-ipex_llm/libs/starcoder.dll,sha256=Bc6UwLdndEvK6Welw7uX9EuhU5TGMz2tvE3FvLurlwY,599040
+ipex_llm/libs/bloom-api.dll,sha256=KHeaZ2ZWeh9XdElXRl_eLX7kmyag077udgNu7l2VDpI,36352
+ipex_llm/libs/bloom.dll,sha256=PNVl7_zJPBqcNY33SNkbl29OBzIwg4Bi99Dea6mqicA,507904
+ipex_llm/libs/gptneox-api.dll,sha256=1cgQlv7rxpf99uAwqxroRgkrBBRSXoveTSYoEkExPQ8,24576
+ipex_llm/libs/gptneox.dll,sha256=BxvNm0dUTlk5Mbbi4QVHAv5wf6lhfoQCp0D0U7qcTPU,568320
+ipex_llm/libs/libbloom_avx.dll,sha256=69_xTl9fesyWpKwWLJ08poZWkltVwMlnRRn3v3ZfGjA,536576
+ipex_llm/libs/libbloom_vnni.dll,sha256=bzty1W9Rz3E2dNU_ukgbh1c3DJrCmxL1_oD5tscZWEw,508416
+ipex_llm/libs/libgptneox_avx.dll,sha256=sRjmFiddbBxmlO6bBN7VkPEEqTQgJVG7Qx0aICzqXYE,596992
+ipex_llm/libs/libgptneox_vnni.dll,sha256=tHuagT6XXNoyZatfmc61tfv4F-m1PnSmS8m5UK3fQnU,568832
+ipex_llm/libs/libllama_avx.dll,sha256=q36jrtKfvTykGX-tq_ppfDJniVyxMTQVSTPat2ugVvQ,591360
+ipex_llm/libs/libllama_vnni.dll,sha256=WFRHJFxMAfHKQNWkjQbBcUCusR2fpC_HNr88qlrNZKs,563200
+ipex_llm/libs/libstarcoder_avx.dll,sha256=hPcOUV5ESsuM0rSS0JybLQMzV8Y9Z4tzq-ZFvAj_B0E,627712
+ipex_llm/libs/libstarcoder_vnni.dll,sha256=idqnOXS2GFu9ImT04POiAEf2mg9-SuH5a55ICcO2OP8,599552
+ipex_llm/libs/llama-api.dll,sha256=NoTN_jIjTreIlIS527X26xXHpO79AIms1fHvn1RJGTg,25600
+ipex_llm/libs/llama.dll,sha256=EtjMujS64UKiemeFKISRWA362TQHZqndTecfa6nArs8,562688
+ipex_llm/libs/main-bloom.exe,sha256=SEnEDBHk5lNErygKppQEAHdASryBqsgm7Qsd-Ymwvn8,103424
+ipex_llm/libs/main-gptneox.exe,sha256=v8041ITCgKMPFEoxKOdTH30Tj4ePkkVff3rwTTlgOQ4,98816
+ipex_llm/libs/main-llama.exe,sha256=-upWWmCLtAMVWXiN1vZz3KnC_UAnHA6pXuNhOZnVPUQ,99840
+ipex_llm/libs/main-starcoder.exe,sha256=hS5dQPdaV323-7Fg6JBENZj0axdgCerivbPUK0rEIgc,157696
+ipex_llm/libs/pipeline.dll,sha256=lB-03EaPL2KqjpZrV29y1YamMrQnLwTFmouYuds7fas,73216
+ipex_llm/libs/quantize-bloom.exe,sha256=XoVeljixMKGiApj1aH2UxdWsRJBUWNxHXklbual33h0,126464
+ipex_llm/libs/quantize-bloom_vnni.exe,sha256=kLMfXy_Vxkno3hI0RD3cpzCTqXEjoiuR35XNkZQpXws,128000
+ipex_llm/libs/quantize-gptneox.exe,sha256=iQMosyaW8RieuC4P-FYiCOB17UwHjpob1cDYo4idJZY,104448
+ipex_llm/libs/quantize-gptneox_vnni.exe,sha256=_rganCVwoTyBDB9Rs-lcU7meSpFq563gZ0vCRiyFqLc,104960
+ipex_llm/libs/quantize-llama.exe,sha256=qGFdhCDOBKKTJV1nreqMFLahjrLoQerJuye8jOjbKC8,110080
+ipex_llm/libs/quantize-llama_vnni.exe,sha256=aQYxCUOj8RYpDujMKLBtLic9OjieET_21oEBMsd0mvY,110592
+ipex_llm/libs/quantize-starcoder.exe,sha256=yEygCz-bQYzY3cOK804M8j87VyhUpDd6pj-fkOJ4WCc,127488
+ipex_llm/libs/quantize-starcoder_vnni.exe,sha256=STkSjhTZ3kIaYobkwglpJg5PH069bfNYBhKcWiOZjQM,128512
+ipex_llm/libs/starcoder-api.dll,sha256=puzt8BGPyAAPrYYmoZ_PcrXfZYDcKW8aJ7w_oKXlrkA,21504
+ipex_llm/libs/starcoder.dll,sha256=JQffgiOYZBfcBqSjMkcJQdG46rGCPed6eIsmU3ucaQ0,599040
 ipex_llm/llamaindex/__init__.py,sha256=T-EbRT6GJ_8RCu-iLmSzcftOimXSPQf2d5X72AUAy2Y,874
 ipex_llm/llamaindex/llms/__init__.py,sha256=KP1lEdGqDuxPoxL1ZSH25Pm2kKMPJBWUTLR0ckSLMIU,1139
 ipex_llm/llamaindex/llms/bigdlllm.py,sha256=FQBzq1KOjfc6uofTXAha3O7TqpJkNfOFepXQmOVlbnI,26314
@@ -87,7 +87,7 @@ ipex_llm/serving/fastchat/tgi_api_protocol.py,sha256=brT3k3-V0NJrU4fRqUwWjC0O3iO
 ipex_llm/serving/fastchat/tgi_api_server.py,sha256=agNTAEiZPSuj3dEdIdYKwkoY0cXOUDX06DiM9VP2knQ,24418
 ipex_llm/serving/fastchat/vllm_worker.py,sha256=ZLz2Q9GxJO6r_LOiP6epgCRjBGk-K4EB1SNEWSJp5DA,11091
 ipex_llm/transformers/__init__.py,sha256=BreA3EY6hkNq0rVixb_sUuTLzMrcWXTt3yRsshCPHHQ,1214
-ipex_llm/transformers/convert.py,sha256=xqqZFGcdDRko2IYgfSgDRs8ef4THUR25IAhSyDV0VUs,106933
+ipex_llm/transformers/convert.py,sha256=niCLZiEQbiMz8UkfYDNwYl2hE1caj93FUR1qjgVh9ZM,107538
 ipex_llm/transformers/convert_ipex.py,sha256=_nSnUTQy-yfkKaqGdqnBdWztZf3NGmnbZ0TKaDrF4X4,14617
 ipex_llm/transformers/embedding.py,sha256=bdgk59DvD4ZZyxRzewXOR7g56nThgO6uhIwk8QL7f-s,9299
 ipex_llm/transformers/kv.py,sha256=src_HcVDKFwQ1V8hdTrFQw5RIwUewM9VOR47GVTPJG4,21187
@@ -138,11 +138,11 @@ ipex_llm/transformers/gguf/models/model_implement/yuan2/configuration_yuan.py,sh
 ipex_llm/transformers/gguf/models/model_implement/yuan2/yuan_hf_model.py,sha256=_AOGMV65XHxgTxIib7lgs49InopcecTzRwgtYR8NTUg,51084
 ipex_llm/transformers/models/__init__.py,sha256=tp2DcVkKg1-QvdYk7DY7rZvQWCDQ4ZjU8NAQ7Fclrpg,584
 ipex_llm/transformers/models/aquila.py,sha256=VZb5Drpo_fTxwcExZ397LygnsNPX2sVbie9_JeFudZI,5252
-ipex_llm/transformers/models/baichuan.py,sha256=8b43mBRZJEf_xLNoodhA4r9x1anqwC3Wt8awWel-aUo,18306
+ipex_llm/transformers/models/baichuan.py,sha256=nRqg_ih8TCIAeSjoiQpaGNAXqXGpVvQmqj5_6gr-Rxg,18395
 ipex_llm/transformers/models/baichuan_m1.py,sha256=l6BD9jbA8TQ5Q0fcngHNRw73SxqvC126ErENs7Zh-do,9513
 ipex_llm/transformers/models/bert.py,sha256=U3JVnuY1z0ILlJgMvetOKuCAkH74ufGbg4whZ8GNtmQ,5620
 ipex_llm/transformers/models/bloom.py,sha256=PxfzyYT-nFn3K5rZhTQjmcEjUUzAhUFzxIN4kzRlCuc,8103
-ipex_llm/transformers/models/chatglm.py,sha256=UHai1t2AUtGmF765_eHF8LUMVQzp_oCBx8TJB21WrHk,12597
+ipex_llm/transformers/models/chatglm.py,sha256=DQM63oPIVMMTBQN4O4hPF4WY1aSiTWq4Bt0BPmgrK0k,12687
 ipex_llm/transformers/models/chatglm2.py,sha256=KyAIX7zGVQDQuwwM3QMBNWZbTeMHEzKUIgAryT0voHc,14933
 ipex_llm/transformers/models/chatglm4.py,sha256=QvUehdaCePB3MNHyWg3dneDxmjtBdxYeKUyQUVcsgfM,16886
 ipex_llm/transformers/models/chatglm4v.py,sha256=Ba9Xtzwtzk_rzg5khGqDrlHfJsDwc5YcM5_yPoord7o,13324
@@ -174,7 +174,7 @@ ipex_llm/transformers/models/qwen2.py,sha256=zK-FpUaxEhjD4gZa1ZvArodAilz29T_cpeA
 ipex_llm/transformers/models/qwen2_5_omni.py,sha256=gNnWvGiPntn6GlyIUtCVYCSIMPU6FG9TCS7fhWbwerY,18779
 ipex_llm/transformers/models/qwen2_moe.py,sha256=a0gYo-ngf8SxaEnBdZUJDnPS6Mkn_poDd8xqhx50icI,19516
 ipex_llm/transformers/models/qwen2_vl.py,sha256=G-9e2oN4f5p5IWQ-zsBZuONxTura3BjlgyT2meigbHQ,13579
-ipex_llm/transformers/models/qwen_vl.py,sha256=j7Nzzz2Qvynu9yrCXmoEfERjw43hXof5TbXIs7Ms-oY,17105
+ipex_llm/transformers/models/qwen_vl.py,sha256=lwNwCJNsBvOu1TGNDW_E2IlCjI7XZwHY8qIcZpuNDFc,17187
 ipex_llm/transformers/models/rwkv4.py,sha256=H4KMtxN0JA2ZTXnonHpsUUJ5xULemo-D1Jzl0ri_UY8,6123
 ipex_llm/transformers/models/rwkv5.py,sha256=OkRNj1pCAZg1z2Fw-I0DEnxLEdZyPeRSQ6msrkxLOCs,10710
 ipex_llm/transformers/models/sd.py,sha256=VvHV5u-0k2MgHu3NL9113hPj7DgfxqctuKzEEeNfRDU,5981
@@ -256,11 +256,11 @@ ipex_llm/vllm/xpu/engine/__init__.py,sha256=pY_CpyuZd72fr6s32ejeKHKFW0K4vUU2rzZj
 ipex_llm/vllm/xpu/engine/engine.py,sha256=NvCMbp0X8NVrOqbwm4FTvXOptTRLzu9jQsy37ZHnTk8,9493
 ipex_llm/vllm/xpu/entrypoints/openai/api_server.py,sha256=IjiSze9vzBCAkLu_VwIcJwuO1jyFna7DLrj6aSL7RaQ,35220
 ipex_llm/vllm/xpu/entrypoints/openai/cli_args.py,sha256=hB398yYtKauASRzevctScdbFIjiiSGMAe1bwEuIHrhY,10893
-ipex_llm-2.3.0b20250413.data/scripts/ipex-llm-init.bat,sha256=HPtCYuDYwEatq7dAwOvdfVcHYCpAVdbj75K1qh0vQek,2578
-ipex_llm-2.3.0b20250413.data/scripts/llm-chat.ps1,sha256=6qrs-hGVAV8IKh7Jx8nq_XrnZcjd7qGU5wndArM7Yag,2769
-ipex_llm-2.3.0b20250413.data/scripts/llm-cli.ps1,sha256=3qBtTLs_EjYDnM8YyCpJhzLnGCKTEGssu9UNqfkjVXs,3009
-ipex_llm-2.3.0b20250413.dist-info/METADATA,sha256=BcY-pLNYgB5lBm5wiYwAn7Y8UBeb5vcMLnTJsjK177A,13917
-ipex_llm-2.3.0b20250413.dist-info/WHEEL,sha256=6iYPr8vTHsyDK75jr9X0V3I9wPSVmtwr_8fdATBciGk,98
-ipex_llm-2.3.0b20250413.dist-info/entry_points.txt,sha256=TiUyBB2MRmfF3ko-pyAEzqeBCRnyhu27bNOAsWPp3e8,61
-ipex_llm-2.3.0b20250413.dist-info/top_level.txt,sha256=CGCMHM-SyqUabU4h8RqJ2KTYckQUO3LvIWwmUQ6Qbzw,9
-ipex_llm-2.3.0b20250413.dist-info/RECORD,,
+ipex_llm-2.3.0b20250415.data/scripts/ipex-llm-init.bat,sha256=HPtCYuDYwEatq7dAwOvdfVcHYCpAVdbj75K1qh0vQek,2578
+ipex_llm-2.3.0b20250415.data/scripts/llm-chat.ps1,sha256=6qrs-hGVAV8IKh7Jx8nq_XrnZcjd7qGU5wndArM7Yag,2769
+ipex_llm-2.3.0b20250415.data/scripts/llm-cli.ps1,sha256=3qBtTLs_EjYDnM8YyCpJhzLnGCKTEGssu9UNqfkjVXs,3009
+ipex_llm-2.3.0b20250415.dist-info/METADATA,sha256=AK6V5GUxadLFvaeVQf6Ym6bSCFwx4ChknK-6c21T2lo,13917
+ipex_llm-2.3.0b20250415.dist-info/WHEEL,sha256=6iYPr8vTHsyDK75jr9X0V3I9wPSVmtwr_8fdATBciGk,98
+ipex_llm-2.3.0b20250415.dist-info/entry_points.txt,sha256=TiUyBB2MRmfF3ko-pyAEzqeBCRnyhu27bNOAsWPp3e8,61
+ipex_llm-2.3.0b20250415.dist-info/top_level.txt,sha256=CGCMHM-SyqUabU4h8RqJ2KTYckQUO3LvIWwmUQ6Qbzw,9
+ipex_llm-2.3.0b20250415.dist-info/RECORD,,