You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Sep 10, 2025. It is now read-only.
<img src="https://pytorch.org/executorch/main/_static/img/llama_ios_app.png" width="600" alt="iOS app running a LlaMA model">
407
408
</a>
409
+
</p>
408
410
</details>
409
411
410
412
@@ -446,15 +448,13 @@ The following assumes you've completed the steps for [Setting up ExecuTorch](#se
446
448
447
449
6. Follow the app's UI guidelines to pick the model and tokenizer files from the local filesystem. Then issue a prompt.
448
450
449
-
<img src="https://pytorch.org/executorch/main/_static/img/android_llama_app.png" width="600" alt="Android app running a LlaMA model">
450
-
451
-
**Note:** The AAR file listed above comes with tiktoken tokenizer, which is used for llama3 model. If you want to use a model with BPE tokenizer (llama2 model for example),
452
-
use this AAR
451
+
**Note:** The AAR file listed in Step 1 has the tiktoken tokenizer, which is used for Llama 3. To tweak or use a custom tokenizer and runtime, modify the ExecuTorch code
452
+
and use [this script](https://github.com/pytorch/executorch/blob/main/build/build_android_llm_demo.sh) to build the AAR library.
* Since the tokenizer is built at compile time, to use a different tokenizer you need to re-build the app.
454
+
<p align="center">
455
+
<img src="https://pytorch.org/executorch/main/_static/img/android_llama_app.png" width="600" alt="Android app running a LlaMA model">
456
+
</p>
456
457
457
-
If you need to tweak or use your own tokenizer and runtime, modify the ExecuTorch code and use [this script](https://github.com/pytorch/executorch/blob/main/build/build_android_llm_demo.sh) to build the AAR library.
0 commit comments