Skip to content

cohere_tokenizer

CohereTokenizer

Bases: BaseTokenizer

Source code in griptape/tokenizers/cohere_tokenizer.py 
13
14
15
16
17
18
19
20
21
@define()
class CohereTokenizer(BaseTokenizer):
    MODEL_PREFIXES_TO_MAX_INPUT_TOKENS = {"command-r": 128000, "command": 4096, "embed": 512}
    MODEL_PREFIXES_TO_MAX_OUTPUT_TOKENS = {"command": 4096, "embed": 512}

    client: Client = field(kw_only=True)

    def count_tokens(self, text: str) -> int:
        return len(self.client.tokenize(text=text, model=self.model).tokens)

MODEL_PREFIXES_TO_MAX_INPUT_TOKENS = {'command-r': 128000, 'command': 4096, 'embed': 512} class-attribute instance-attribute

MODEL_PREFIXES_TO_MAX_OUTPUT_TOKENS = {'command': 4096, 'embed': 512} class-attribute instance-attribute

client = field(kw_only=True) class-attribute instance-attribute

count_tokens(text)

Source code in griptape/tokenizers/cohere_tokenizer.py 
20
21
def count_tokens(self, text: str) -> int:
    return len(self.client.tokenize(text=text, model=self.model).tokens)