Spaces:
Sleeping
Sleeping
File size: 2,206 Bytes
1b74e0a 6a286a9 1b74e0a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 |
from transformers import T5EncoderModel, T5Config
from huggingface_hub import hf_hub_download
import torch.nn as nn
import torch
NUM_LABELS = 4
class T5ClassificationModel(nn.Module):
def __init__(self, model_path="t5-small", freeze_weights=True):
super(T5ClassificationModel, self).__init__()
if model_path == "t5-small":
self.base_model = T5EncoderModel.from_pretrained(model_path)
else:
pytorch_model_path = hf_hub_download(
repo_id=model_path,
repo_type="model",
filename="pytorch_model.bin"
)
config = T5Config.from_pretrained(model_path)
self.base_model = T5EncoderModel(config)
# Load the state_dict and remove unwanted keys
state_dict = torch.load(pytorch_model_path, map_location=torch.device("cpu"))
filtered_state_dict = {
k.replace("base_model.", ""): v
for k, v in state_dict.items()
if not k.startswith("classifier.")
}
self.base_model.load_state_dict(filtered_state_dict)
# For push to hub.
self.config = self.base_model.config
# Freeze the base model's weights
if freeze_weights:
for param in self.base_model.parameters():
param.requires_grad = False
# Add a classification head
self.classifier = nn.Linear(self.base_model.config.hidden_size, NUM_LABELS)
def forward(self, input_ids, attention_mask, labels=None):
with torch.no_grad(): # No gradients for the base model
outputs = self.base_model(input_ids=input_ids, attention_mask=attention_mask)
# Sum token representations
summed_representation = outputs.last_hidden_state.sum(dim=1) # Summing over the sequence length (dim=1)
logits = self.classifier(summed_representation) # Pass the summed representation to the classifier
loss = None
if labels is not None:
loss_fn = nn.BCEWithLogitsLoss()
loss = loss_fn(logits, labels.float())
return {"loss": loss, "logits": logits}
|