ó
Ù^íh' ã ób • S SK JrJrJr S SKrS SKJr " S S\5 r\S r/ SQr S\
4S jr g) é )ÚLiteralÚOptionalÚUnionN)ÚBaseLLMExceptionc
ób ^ • \ rS rSr SS\S\S\\\\ R 4 4U 4S jjjr Sr U =r
$ )ÚHuggingfaceErroré Ú status_codeÚmessageÚheadersc ó" >• [ TU ] XUS9 g )N)r
r r )ÚsuperÚ__init__)Úselfr
r r Ú __class__s €Ú_/home/james-whalen/.local/lib/python3.13/site-packages/litellm/llms/huggingface/common_utils.pyr ÚHuggingfaceError.__init__ s ø€ ô ‰Ñ [È7ÐÒSó © )N)Ú__name__Ú
__module__Ú __qualname__Ú__firstlineno__ÚintÚstrr r ÚdictÚhttpxÚHeadersr Ú__static_attributes__Ú
__classcell__)r s @r r r sF ø† ð
9=ñ TàðTð ðTð ˜% e§m¡mÐ 3Ñ4Ñ5÷ Tö Tr r )ztext-generation-inferenceÚconversationalztext-classificationztext-generationÚgenerated_textc ó • / SQnU Hw nU R 5 R U5 ( a U R USS5 n U R U5 ( d MR U SSS2 R USSS2 SS5 SSS2 n My U $ )zÄ
Parse the output text to remove any special characters. In our current approach we just check for ChatML tokens.
Initial issue that prompted this - https://github.com/BerriAI/litellm/issues/763
)z
<|assistant|>z
<|system|>z<|user|>zzÚ é Néÿÿÿÿ)ÚstripÚ
startswithÚreplaceÚendswith)r" Úchat_template_tokensÚtokens r Ú
output_parserr- ! s‹ € ò VÐÛ%ˆØ × Ñ Ó !× ,Ñ ,¨U× 3Ñ 3Ø +× 3Ñ 3°E¸2¸qÓ AˆNØ × "Ñ " 5× )Ó )Ø +©D¨b¨DÑ 1× 9Ñ 9¸%ÁÀ"À¹+ÀrÈ1Ó MÉdÐPRÈdÑ SŠNñ &ð
Ðr ) Útypingr r r r Ú)litellm.llms.base_llm.chat.transformationr r Úhf_tasksÚ hf_task_listr r- r r r Úr2 sD ðß +Ñ +ã å FôTÐ'ô Tð ðñ €ò€ ð #õ r