The testing model from HF doesn't have the tokenizer max length set c…

…orrectly

stanza/models/common/bert_embedding.py

@@ -24,7 +24,8 @@ def __init__(self, length, max_len, line_num, text):
 
 
 def update_max_length(model_name, tokenizer):
-    if model_name in ('google/muril-base-cased',
+    if model_name in ('hf-internal-testing/tiny-bert',
+                      'google/muril-base-cased',
                       'google/muril-large-cased',
                       'airesearch/wangchanberta-base-att-spm-uncased',
                       'camembert/camembert-large',