Update DeepSeek model

.gitattributes

@@ -34,3 +34,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 deepseek_q8_ekv1280.task filter=lfs diff=lfs merge=lfs -text
+DeepSeek-R1-Distill-Qwen-1.5B_multi-prefill-seq_f32_ekv1280.task filter=lfs diff=lfs merge=lfs -text
+DeepSeek-R1-Distill-Qwen-1.5B_multi-prefill-seq_q8_ekv1280.task filter=lfs diff=lfs merge=lfs -text

DeepSeek-R1-Distill-Qwen-1.5B_multi-prefill-seq_f32_ekv1280.task

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:10dd74836637508364b953448f1222981deaafa8ad0af2f7dbc683794b4e84cf
+size 7124558842

DeepSeek-R1-Distill-Qwen-1.5B_multi-prefill-seq_f32_ekv1280.tflite

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4eb5bae61d9717fd9d0ff9c1f00266e27ab049ec807b1e9dd1440041600bcdfc
+size 7121909824

DeepSeek-R1-Distill-Qwen-1.5B_multi-prefill-seq_q8_ekv1280.task

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b1311ae8c3ad18d6c71abdf6859f63908aebe20bf42b18ba511c5fa033c179ee
+size 1861094766

DeepSeek-R1-Distill-Qwen-1.5B_multi-prefill-seq_q8_ekv1280.tflite

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ad356efd69259876152347e51cf776443a4fce3abe61d7212bb685c702d70560
+size 1858445888

DeepSeek-R1-Distill-Qwen-1.5B_seq128_f32_ekv1280.tflite

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba4dd48eec47d612b9501f8365671e4a9714487a67c4276a00309637e99f7a02
+size 7116878944

DeepSeek-R1-Distill-Qwen-1.5B_seq128_q8_ekv1280.tflite

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7b06368fa46eb6934daf035e5bcc7bdd37086f0fd92fbbe9ce744613a379e209
+size 1806773448

README.md

@@ -1,34 +1,45 @@
----
+--------------------------------------------------------------------------------
+
 license: mit
 base_model:
 - deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
----
 
 # litert-community/DeepSeek-R1-Distill-Qwen-1.5B
 
-This model provides a few variants of [deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B) that are ready for deployment on Android using the [LiteRT (fka TFLite) stack](https://ai.google.dev/edge/litert) and [MediaPipe LLM Inference API](https://ai.google.dev/edge/mediapipe/solutions/genai/llm_inference).
+This model provides a few variants of
+[deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B) that are ready for
+deployment on Android using the
+[LiteRT (fka TFLite) stack](https://ai.google.dev/edge/litert) and
+[MediaPipe LLM Inference API](https://ai.google.dev/edge/mediapipe/solutions/genai/llm_inference).
 
 ## Use the models
 
 ### Colab
 
-*Disclaimer: The target deployment surface for the LiteRT models is Android/iOS/Web and the stack has been optimized for performance on these targets. Trying out the system in Colab is an easier way to familiarize yourself with the LiteRT stack, with the caveat that the performance (memory and latency) on Colab could be much worse than on a local device.*
+*Disclaimer: The target deployment surface for the LiteRT models is
+Android/iOS/Web and the stack has been optimized for performance on these
+targets. Trying out the system in Colab is an easier way to familiarize yourself
+with the LiteRT stack, with the caveat that the performance (memory and latency)
+on Colab could be much worse than on a local device.*
 
-[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/#fileId=https://huggingface.co/litert-community/DeepSeek-R1-Distill-Qwen-1.5B/blob/main/deepseek_tflite.ipynb)
+[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/#fileId=https://huggingface.co/litert-community/DeepSeek-R1-Distill-Qwen-1.5B/blob/main/DeepSeek-R1-Distill-Qwen-1.5B_tflite.ipynb)
 
 ### Android
 
-* Download and install [the apk](https://github.com/google-ai-edge/mediapipe-samples/releases/download/v0.1.0/llm_inference_v0.1.0-debug.apk).
-* Follow the instructions in the app.
-
+*   Download and install
+    [the apk](https://github.com/google-ai-edge/mediapipe-samples/releases/latest/download/llm_inference-debug.apk).
+*   Follow the instructions in the app.
 
-To build the demo app from source, please follow the [instructions](https://github.com/google-ai-edge/mediapipe-samples/blob/main/examples/llm_inference/android/README.md) from the GitHub repository.
+To build the demo app from source, please follow the
+[instructions](https://github.com/google-ai-edge/mediapipe-samples/blob/main/examples/llm_inference/android/README.md)
+from the GitHub repository.
 
 ## Performance
 
 ### Android
 
-Note that all benchmark stats are from a Samsung S24 Ultra with 1280 KV cache size, 512 tokens prefill, 128 tokens decode.
+Note that all benchmark stats are from a Samsung S24 Ultra with 
+1280 KV cache size with multiple prefill signatures enabled.
 
 <table border="1">
   <tr>
@@ -41,26 +52,30 @@ Note that all benchmark stats are from a Samsung S24 Ultra with 1280 KV cache si
    <th>Model size (MB)</th>
   </tr>
   <tr>
-   <td>fp32 (baseline)</td>
-   <td rowspan="2">CPU</td>
-   <td><p style="text-align: right">45</p></td>
-   <td><p style="text-align: right">6</p></td>
-   <td><p style="text-align: right">8</p></td>
-   <td><p style="text-align: right">6,213</p></td>
-   <td><p style="text-align: right">7,124</p></td>
-  </tr>
-  <tr>
-   <td>dynamic_int8</td>
-   <td><p style="text-align: right">261</p></td>
-   <td><p style="text-align: right">23</p></td>
-   <td><p style="text-align: right">2 </p></td>
-   <td><p style="text-align: right">1,936 </p></td>
-   <td><p style="text-align: right">1,861</p></td>
-  </tr>
+<td>fp32 (baseline)</td>
+<td>cpu</td>
+<td><p style="text-align: right">41.84 tk/s</p></td>
+<td><p style="text-align: right">6.14 tk/s</p></td>
+<td><p style="text-align: right">14.30 s</p></td>
+<td><p style="text-align: right">7,421 MB</p></td>
+<td><p style="text-align: right">6,794 MB</p></td>
+</tr>
+<tr>
+<td>dynamic_int8</td>
+<td>cpu</td>
+<td><p style="text-align: right">228.57 tk/s</p></td>
+<td><p style="text-align: right">18.80 tk/s</p></td>
+<td><p style="text-align: right">3.14 s</p></td>
+<td><p style="text-align: right">3,600 MB</p></td>
+<td><p style="text-align: right">1,774 MB</p></td>
+</tr>
+
 </table>
 
-*   Model Size: measured by the size of the .tflite flatbuffer (serialization format for LiteRT models)
+*   Model Size: measured by the size of the .tflite flatbuffer (serialization
+    format for LiteRT models)
 *   Memory: indicator of peak RAM usage
-*   The inference on CPU is accelerated via the LiteRT [XNNPACK](https://github.com/google/XNNPACK) delegate with 4 threads
+*   The inference on CPU is accelerated via the LiteRT
+    [XNNPACK](https://github.com/google/XNNPACK) delegate with 4 threads
 *   Benchmark is done assuming XNNPACK cache is enabled
-*   dynamic_int8: quantized model with int8 weights and float activations.
+*   dynamic_int8: quantized model with int8 weights and float activations.

tokenizer.model

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5baf9d2f6c82b4e4eeeb7147babd3d82682381f5ee83a78faee472294dce457b
+size 2648396