OAI: Add response_prefix and fix BOS token issues in chat completions

response_prefix is used to add a prefix before generating the next
message. This is used in many cases such as continuining a prompt
(see #96).

Also if a template has BOS token specified, add_bos_token will
append two BOS tokens. Add a check which strips a starting BOS token
from the prompt if it exists.

Signed-off-by: kingbri <bdashore3@proton.me>

backends/exllamav2/model.py

@@ -878,6 +878,7 @@ class ExllamaV2Container:
             encode_special_tokens=True,
             return_offsets=True,
         )
+        print(ids)
         mask = (
             self.tokenizer.padding_mask(ids)
             if self.use_cfg and gen_settings.cfg_scale not in [None, 1.0]