πŸŒ™ Jungian Dream Archetype Classifier β€” group_3

Fine-tuned bert-base-uncased for multi-label Jungian archetype detection in dream narratives.

Archetypes in this model

  • The_Child
  • The_Persona
  • The_Lover
  • The_Magician

Model details

Base model bert-base-uncased
Task Multi-label classification
Max length 256 tokens
Labels 4 archetypes

Usage

import torch, json
from transformers import BertTokenizer, BertForSequenceClassification
from huggingface_hub import hf_hub_download

REPO = "Aniruddhhhh/jungian-dream-archetype-group_3"
tokenizer = BertTokenizer.from_pretrained(REPO)
model     = BertForSequenceClassification.from_pretrained(REPO)
model.eval()

arch_path  = hf_hub_download(REPO, "archetypes.json")
with open(arch_path) as f:
    archetypes = json.load(f)

dream = "I was chased through a dark forest by a shadowy figure."
enc   = tokenizer(dream, return_tensors="pt",
                  truncation=True, max_length=256)
with torch.no_grad():
    probs = torch.sigmoid(model(**enc).logits)[0]
for arch, prob in zip(archetypes, probs):
    print(f"{arch}: {prob:.3f}")
Downloads last month
7
Safetensors
Model size
0.1B params
Tensor type
F32
Β·
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support