[Bug] Failed to load Llama

[mariagcomesanaia-spa]

🐛 Bug

To Reproduce

Steps to reproduce the behavior:

1.Load LLama model
1.
1.

Expected behavior

response

MLCChat failed

Stack trace:
org.apache.tvm.Base$TVMError: InternalError: Check failed: (config["conv_template"].isstd::string()) is false:
Stack trace:
File "/Users/kartik/mlc/mlc-llm/cpp/llm_chat.cc", line 540

at org.apache.tvm.Base.checkCall(Base.java:173)
at org.apache.tvm.Function.invoke(Function.java:130)
at ai.mlc.mlcllm.ChatModule.reload(ChatModule.java:46)
at ai.mlc.mlcchat.AppViewModel$ChatState$mainReloadChat$1$2.invoke(AppViewModel.kt:648)
at ai.mlc.mlcchat.AppViewModel$ChatState$mainReloadChat$1$2.invoke(AppViewModel.kt:646)
at ai.mlc.mlcchat.AppViewModel$ChatState.callBackend(AppViewModel.kt:548)
at ai.mlc.mlcchat.AppViewModel$ChatState.mainReloadChat$lambda$3(AppViewModel.kt:646)
at ai.mlc.mlcchat.AppViewModel$ChatState.$r8$lambda$CXL6v4mjTu_Sr5Pk2zFDcus0R-8(Unknown Source:0)
at ai.mlc.mlcchat.AppViewModel$ChatState$$ExternalSyntheticLambda2.run(Unknown Source:8)
at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:524)
at java.util.concurrent.FutureTask.run(FutureTask.java:317)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1156)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:651)
at java.lang.Thread.run(Thread.java:1119)

Error message:
InternalError: Check failed: (config["conv_template"].isstd::string()) is false:
Stack trace:
File "/Users/kartik/mlc/mlc-llm/cpp/llm_chat.cc", line 540

Environment

• Platform (e.g. WebGPU/Vulkan/IOS/Android/CUDA): Android 15
• Operating system (e.g. Ubuntu/Windows/MacOS/...):
• Device (e.g. iPhone 12 Pro, PC+RTX 3090, ...) Xiaoimi Note 13 pro+
• How you installed MLC-LLM (conda, source):
• How you installed TVM (pip, source):
• Python version (e.g. 3.10):
• GPU driver version (if applicable):
• CUDA/cuDNN version (if applicable):
• TVM Hash Tag (python -c "import tvm; print('\n'.join(f'{k}: {v}' for k, v in tvm.support.libinfo().items()))", applicable if you compile models):
• Any other relevant information:

Additional context

[MasterJH5574]

Thank you @mariagcomesanaia-spa for reporting. Do you mind sharing the concrete model HuggingFace repo link? So we can try to reproduce the issue.