Merge branch 'fix_bncmark_bug' into 'dev-rocm'

fix param s=1 benchmark crash bug

See merge request huangwb1/text-generation-inference!1

server/text_generation_server/models/flash_causal_lm.py

@@ -169,11 +169,13 @@ class FlashCausalLMBatch(Batch):
             requests_idx_mapping[r.id] = i
 
             tokenized_input = tokenized_input[-r.truncate :]
+            '''fix input s=1 crash bug
             if (
                 tokenized_input[0] == tokenizer.bos_token_id
                 and tokenized_input[1] == tokenizer.bos_token_id
             ):
                 tokenized_input = tokenized_input[1:]
+            '''
 
             input_length = len(tokenized_input)
             input_lengths.append(input_length)