From 19a43322a9b1c4e8edfbca6824ce5f2084c75fc3 Mon Sep 17 00:00:00 2001
From: anakinxc <103552181+anakinxc@users.noreply.github.com>
Date: Tue, 19 Dec 2023 18:07:52 +0800
Subject: [PATCH] Update README.md

---
 examples/python/ml/flax_llama7b/README.md | 22 ----------------------
 1 file changed, 22 deletions(-)

diff --git a/examples/python/ml/flax_llama7b/README.md b/examples/python/ml/flax_llama7b/README.md
index 949947dc..ab1f5a90 100644
--- a/examples/python/ml/flax_llama7b/README.md
+++ b/examples/python/ml/flax_llama7b/README.md
@@ -33,28 +33,6 @@ This example demonstrates how to use SPU to run secure inference on a pre-traine
        parser.add_argument("--streaming", action="store_true", default=False, help="whether is model weight saved stream format",)
     ```
 
-    Since EasyLM have an issue，so we have to make a samll change to support the option "streaming=false".
-    Open and edit "convert_hf_to_easylm.py", chang this:
-
-    ```python
-       parser.add_argument("--streaming", action="store_true", default=True, help="whether is model weight saved stream format",)
-    ```
-
-    to:
-
-    ```python
-       parser.add_argument("--streaming", action="store_true", default=False, help="whether is model weight saved stream format",)
-    ```
-
-    Since EasyLM have an issue，so we have to make a samll change to support the option "streaming=false".
-    Open and edit "convert_hf_to_easylm.py", chang this:
-    ```python
-       parser.add_argument("--streaming", action="store_true", default=True, help="whether is model weight saved stream format",)
-    ```
-    to:
-    ```python
-       parser.add_argument("--streaming", action="store_true", default=False, help="whether is model weight saved stream format",)
-    ```
     Download trained LLaMA-B[PyTroch-Version] from [Hugging Face](https://huggingface.co/openlm-research/open_llama_7b)
     , and convert it to Flax.msgpack as:
     ```sh