Commit 64ffeed1 authored by dmMaze's avatar dmMaze
Browse files

try to fix chatgpt (#173)

parent 3006d2a4

modules/translators/trans_chatgpt.py

+26 −12
Original line number Diff line number Diff line
@@ -8,7 +8,9 @@ from typing import List, Dict, Union 
import yaml

from .base import BaseTranslator, register_translator



CONFIG = None



class InvalidNumTranslations(Exception):

    pass



@register_translator('chatgpt')

class GPTTranslator(BaseTranslator):
@@ -23,7 +25,7 @@ class GPTTranslator(BaseTranslator): 
                'gpt35-turbo',

                'gpt4',

            ],

            'select': 'gpt3'

            'select': 'gpt35-turbo'

        },

        'prompt template': {

            'type': 'editor',
@@ -142,12 +144,14 @@ class GPTTranslator(BaseTranslator): 
        prompt = ''

        max_tokens = self.max_tokens

        # return_prompt = self.params['return prompt']

        prompt_template = self.params['prompt template']['content'].format(to_lang=to_lang)

        prompt_template = self.params['prompt template']['content'].format(to_lang=to_lang).rstrip()

        prompt += prompt_template



        i_offset = 0

        num_src = 0

        for i, query in enumerate(queries):

            prompt += f'\n<|{i+1-i_offset}|>{query}'

            num_src += 1

            # If prompt is growing too large and theres still a lot of text left

            # split off the rest of the queries into new prompts.

            # 1 token = ~4 characters according to https://platform.openai.com/tokenizer
@@ -155,14 +159,15 @@ class GPTTranslator(BaseTranslator): 
            if max_tokens * 2 and len(''.join(queries[i+1:])) > max_tokens:

                # if return_prompt:

                #     prompt += '\n<|1|>'

                yield prompt.lstrip()

                yield prompt.lstrip(), num_src

                prompt = prompt_template

                # Restart counting at 1

                i_offset = i + 1

                num_src = 0



        # if return_prompt:

        #     prompt += '\n<|1|>'

        yield prompt.lstrip()

        yield prompt.lstrip(), num_src



    def _format_prompt_log(self, to_lang: str, prompt: str) -> str:

        chat_sample = self.chat_sample
@@ -193,12 +198,15 @@ class GPTTranslator(BaseTranslator): 
        queries = text

        # return_prompt = self.params['return prompt']

        chat_sample = self.chat_sample

        for prompt in self._assemble_prompts(from_lang, to_lang, queries):

        for prompt, num_src in self._assemble_prompts(from_lang, to_lang, queries):

            ratelimit_attempt = 0

            server_error_attempt = 0

            retry_attempt = 0

            while True:

                try:

                    response = self._request_translation(prompt, chat_sample)

                    new_translations = re.split(r'<\|\d+\|>', response)[-num_src:]

                    if len(new_translations) != num_src:

                        raise InvalidNumTranslations

                    break

                except openai.error.RateLimitError: # Server returned ratelimit response

                    ratelimit_attempt += 1
@@ -207,15 +215,21 @@ class GPTTranslator(BaseTranslator): 
                    self.logger.warn(f'Restarting request due to ratelimiting by openai servers. Attempt: {ratelimit_attempt}')

                    time.sleep(2)

                except openai.error.APIError: # Server returned 500 error (probably server load)

                    server_error_attempt += 1

                    if server_error_attempt >= self.retry_attempts:

                    retry_attempt += 1

                    if retry_attempt >= self.retry_attempts:

                        self.logger.error('OpenAI encountered a server error, possibly due to high server load. Use a different translator or try again later.')

                        raise

                    self.logger.warn(f'Restarting request due to a server error. Attempt: {server_error_attempt}')

                    self.logger.warn(f'Restarting request due to a server error. Attempt: {retry_attempt}')

                    time.sleep(1)

                except InvalidNumTranslations:

                    retry_attempt += 1

                    message = f'number of translations does not match to source:\nprompt:\n    {prompt}\ntranslations:\n  {new_translations}\nopenai response:\n  {response}'

                    if retry_attempt >= self.retry_attempts:

                        self.logger.error(message)

                        new_translations = [''] * num_src

                        break

                    self.logger.warn(message + '\n' + f'Restarting request. Attempt: {retry_attempt}')



            self.logger.debug('-- GPT Response --\n' + response)

            new_translations = re.split(r'<\|\d+\|>', response)[-len(queries):]

            # if return_prompt:

            #     new_translations = new_translations[:-1]