haznitrama/babybabellm-multi_gpu-gpt_bert-eng-classifier_head

GPT-BERT style BabyBabyLLM model for language eng.

This repository may include both main and EMA variants.

Default variant exposed to generic loaders: main

Variants Available

main

Files

• model.safetensors (alias of default variant)

Configuration

{
  "attention_probs_dropout_prob": 0.1,
  "hidden_dropout_prob": 0.1,
  "hidden_size": 768,
  "intermediate_size": 2560,
  "max_position_embeddings": 512,
  "position_bucket_size": 32,
  "num_attention_heads": 12,
  "num_hidden_layers": 12,
  "vocab_size": 16384,
  "layer_norm_eps": 1e-05,
  "force_causal_mask": true,
  "classifier_dropout": 0.1,
  "classifier_layer_norm_eps": 1e-05,
  "num_labels": 2
}

Tokenizer file: tokenizer_eng.json

Quick Usage

from transformers import AutoTokenizer, AutoModelForMaskedLM
model_id = 'haznitrama/babybabellm-multi_gpu-gpt_bert-eng-classifier_head'
tok = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForMaskedLM.from_pretrained(model_id, trust_remote_code=True)
out = model(**tok('Hello world', return_tensors='pt'))

Forced Causal Attention

Causal attention is enforced during inference by applying a triangular future mask inside the remote code. This prevents the hybrid GPT-BERT layers from attending to future tokens even when a bidirectional mask is provided.

Sequence Classification

GPTBertForSequenceClassification mirrors the original GLUE classifier head for downstream fine-tuning.

from transformers import AutoTokenizer, AutoModelForSequenceClassification
model_id = 'haznitrama/babybabellm-multi_gpu-gpt_bert-eng-classifier_head'
tok = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForSequenceClassification.from_pretrained(model_id, trust_remote_code=True)
outputs = model(**tok('This movie was great!', return_tensors='pt'))
print(outputs.logits)

Notes

• Converted on 2025-10-03T22:11:01.747086+00:00
• Weights are the exact trained parameters; no new layers were initialized.
• Requires trust_remote_code=True due to custom architecture.