Quantumbraid's picture
download
raw
400 Bytes
from transformers import LlamaForCausalLM
from .._mixin import OstrisTransformersMixin
class LlamaTextEncoder(LlamaForCausalLM, OstrisTransformersMixin):
"""Llama causal-LM text encoder (hidream's text_encoder_4)."""
aitk_subfolder = "text_encoder_4"
aitk_tokenizer_subfolder = "tokenizer_4"
@classmethod
def get_transformer_block_names(cls):
return ["model.layers"]

Xet Storage Details

Size:
400 Bytes
·
Xet hash:
d262fef3ed3c1a610d80d83be10d2873d32f74875454cc6ecec1b5b628e98075

Xet efficiently stores files, intelligently splitting them into unique chunks and accelerating uploads and downloads. More info.