From d02fcf192ac10e00e9dacffb820c1f460dae9c6a Mon Sep 17 00:00:00 2001
From: Chen Lai <chenlai@meta.com>
Date: Thu, 4 Sep 2025 15:46:33 -0700
Subject: [PATCH] Support qwen 0.5 b (#13972)

Summary:

Add entry point to run 0.5B model. Run with
```
 buck run mode/dev-nosan //executorch/extension/llm/export:export_llm -- --config executorch/examples/models/llama/config/llama_bf16.yaml +base.model_class="qwen2_5_0_5b" +base.params="./executorch/examples/models/qwen2_5/config/0_5b_config.json"
```

Differential Revision: D81724880
---
 examples/models/llama/export_llama_lib.py | 8 +++++---
 examples/models/qwen2_5/README.md         | 7 ++++---
 extension/llm/export/config/llm_config.py | 3 ++-
 3 files changed, 11 insertions(+), 7 deletions(-)

diff --git a/examples/models/llama/export_llama_lib.py b/examples/models/llama/export_llama_lib.py
index 07de9f3fa75..d0e45553957 100644
--- a/examples/models/llama/export_llama_lib.py
+++ b/examples/models/llama/export_llama_lib.py
@@ -93,7 +93,8 @@
     "llama3_1",
     "llama3_2",
     "static_llama",
-    "qwen2_5",
+    "qwen2_5_0_5b",
+    "qwen2_5_1_5b",
     "qwen3_0_6b",
     "qwen3_1_7b",
     "qwen3_4b",
@@ -102,7 +103,8 @@
 ]
 TORCHTUNE_DEFINED_MODELS = ["llama3_2_vision"]
 HUGGING_FACE_REPO_IDS = {
-    "qwen2_5": "Qwen/Qwen2.5-1.5B",
+    "qwen2_5_0_5b": "Qwen/Qwen2.5-0.5B",
+    "qwen2_5_1_5b": "Qwen/Qwen2.5-1.5B",
     "phi_4_mini": "microsoft/Phi-4-mini-instruct",
     "smollm2": "HuggingFaceTB/SmolLM-135M",
     "qwen3_0_6b": "Qwen/Qwen3-0.6B",
@@ -595,7 +597,7 @@ def export_llama(
     model_name = llm_config.base.model_class.value
     if not llm_config.base.checkpoint and model_name in HUGGING_FACE_REPO_IDS:
         repo_id = HUGGING_FACE_REPO_IDS[model_name]
-        if model_name == "qwen2_5":
+        if model_name.startswith("qwen2_5"):
             from executorch.examples.models.qwen2_5 import convert_weights
         elif model_name.startswith("qwen3"):
             from executorch.examples.models.qwen3 import convert_weights
diff --git a/examples/models/qwen2_5/README.md b/examples/models/qwen2_5/README.md
index c58807b46cb..11c92f4bd73 100644
--- a/examples/models/qwen2_5/README.md
+++ b/examples/models/qwen2_5/README.md
@@ -6,9 +6,10 @@ Qwen 2.5 is the latest iteration of the Qwen series of large language models (LL
 Qwen 2.5 uses the same example code as Llama, while the checkpoint, model params, and tokenizer are different. Please see the [Llama README page](../llama/README.md) for details.
 
 All commands for exporting and running Llama on various backends should also be applicable to Qwen 2.5, by swapping the following args:
+
 ```
-base.model_class="qwen2_5"
-base.params="examples/models/qwen2_5/config/1_5b_config.json"
+base.model_class=[qwen2_5_0_5b, qwen2_5_1_5b]
+base.params=[examples/models/qwen2_5/config/0_5b_config.json, examples/models/qwen2_5/config/1_5b_config.json]
 base.checkpoint=<path-to-meta-checkpoint>
 ```
 
@@ -34,7 +35,7 @@ QWEN_CHECKPOINT=path/to/checkpoint.pth
 
 python -m extension.llm.export.export_llm \
   --config examples/models/qwen2_5/config/qwen2_5_xnnpack_q8da4w.yaml \
-  +base.model_class="qwen2_5" \
+  +base.model_class="qwen2_5_1_5b" \
   +base.checkpoint="${QWEN_CHECKPOINT:?}" \
   +base.params="examples/models/qwen2_5/config/1_5b_config.json" \
   +export.output_name="qwen2_5-1_5b.pte" \
diff --git a/extension/llm/export/config/llm_config.py b/extension/llm/export/config/llm_config.py
index 8f8646e88cc..839ef7bc730 100644
--- a/extension/llm/export/config/llm_config.py
+++ b/extension/llm/export/config/llm_config.py
@@ -37,7 +37,8 @@ class ModelType(str, Enum):
     llama3_2 = "llama3_2"
     llama3_2_vision = "llama3_2_vision"
     static_llama = "static_llama"
-    qwen2_5 = "qwen2_5"
+    qwen2_5_0_5b = "qwen2_5_0_5b"
+    qwen2_5_1_5b = "qwen2_5_1_5b"
     qwen3_0_6b = "qwen3_0_6b"
     qwen3_1_7b = "qwen3_1_7b"
     qwen3_4b = "qwen3_4b"