update axcl aarch64 bin

.gitattributes

@@ -38,3 +38,4 @@ main_axcl_aarch64 filter=lfs diff=lfs merge=lfs -text
 main_axcl_x86 filter=lfs diff=lfs merge=lfs -text
 main_prefill filter=lfs diff=lfs merge=lfs -text
 main_ax650 filter=lfs diff=lfs merge=lfs -text
+main filter=lfs diff=lfs merge=lfs -text

main

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9396632c6f0ad5dde9f93a5e4ac57cd3cbb3d2fd1b3748cfdd188bbbce0bfe62
+size 1787328

run_smollm2_360m_axcl_aarch64.sh

@@ -1,14 +1,11 @@
-./main_axcl_aarch64 \
+./main \
 --template_filename_axmodel "smollm2-360m-ax650/llama_p128_l%d_together.axmodel" \
 --axmodel_num 32 \
---tokenizer_type 2 \
---filename_tokenizer_model "http://127.0.0.1:12345" \
---bos 0 --eos 0 \
+--url_tokenizer_model "http://127.0.0.1:12345" \
 --filename_post_axmodel "smollm2-360m-ax650/llama_post.axmodel" \
 --filename_tokens_embed "smollm2-360m-ax650/model.embed_tokens.weight.bfloat16.bin" \
 --tokens_embed_num 49152 \
 --tokens_embed_size 960 \
 --use_mmap_load_embed 0 \
 --live_print 1 \
---continue 1 \
---prompt "$1"
+--devices 0

smollm2_tokenizer_uid.py

@@ -11,7 +11,7 @@ class Tokenizer_Http():
     def __init__(self):
         model_id = "smollm2_tokenizer"
         self.tokenizer = AutoTokenizer.from_pretrained(model_id)
-        self.messages = [{"role": "system", "content": "I are a AI assistant, created by HuggingfaceTB"}]
+        self.messages = [{"role": "system", "content": "You are a AI assistant, created by HuggingfaceTB"}]
         self.token_ids = []
 
     def encode(self, prompt, last_reply=None):
@@ -58,7 +58,7 @@ class Tokenizer_Http():
     def eos_token(self):
         return self.tokenizer.eos_token
     
-    def reset(self, system_prompt="I are a AI assistant, created by HuggingfaceTB"):
+    def reset(self, system_prompt="You are a AI assistant, created by HuggingfaceTB"):
         self.messages = [
             {"role": "system", "content": system_prompt},
         ]