0 Fork 0

Files

T

Manuel Romero 6083c1566e Update README.md

I trained the model for more epochs so I improved the results. This commit will update the results of the model and add a gif using it with **transformers/pipelines**

2020-02-16 10:09:34 -05:00

3.6 KiB

Raw Blame History

language, thumbnail

language	thumbnail
spanish	https://i.imgur.com/jgBdimh.png

BETO (Spanish BERT) + Spanish SQuAD2.0 + distillation using 'bert-base-multilingual-cased' as teacher

This model is a fine-tuned on SQuAD-es-v2.0 and distilled version of BETO for Q&A.

Distillation makes the model smaller, faster, cheaper and lighter than bert-base-spanish-wwm-cased-finetuned-spa-squad2-es

This model was fine-tuned on the same dataset but using distillation during the process as mentioned above (and one more train epoch).

The teacher model for the distillation was bert-base-multilingual-cased. It is the same teacher used for distilbert-base-multilingual-cased AKA DistilmBERT (on average is twice as fast as mBERT-base).

Details of the downstream task (Q&A) - Dataset

SQuAD-es-v2.0

Dataset	# Q&A
SQuAD2.0 Train	130 K
SQuAD2.0-es-v2.0	111 K
SQuAD2.0 Dev	12 K
SQuAD-es-v2.0-small Dev	69 K

Model training

The model was trained on a Tesla P100 GPU and 25GB of RAM with the following command:

!export SQUAD_DIR=/path/to/squad-v2_spanish \
&& python transformers/examples/distillation/run_squad_w_distillation.py \
  --model_type bert \
  --model_name_or_path dccuchile/bert-base-spanish-wwm-cased \
  --teacher_type bert \
  --teacher_name_or_path bert-base-multilingual-cased \
  --do_train \
  --do_eval \
  --do_lower_case \
  --train_file $SQUAD_DIR/train-v2.json \
  --predict_file $SQUAD_DIR/dev-v2.json \
  --per_gpu_train_batch_size 12 \
  --learning_rate 3e-5 \
  --num_train_epochs 5.0 \
  --max_seq_length 384 \
  --doc_stride 128 \
  --output_dir /content/model_output \
  --save_steps 5000 \
  --threads 4 \
  --version_2_with_negative

Results:

Metric	# Value
Exact	90.7748
F1	94.9471

{
  "exact": 90.77483309730933,
  "f1": 94.94714391266254,
  "total": 69202,
  "HasAns_exact": 86.60850599781898,
  "HasAns_f1": 92.90582885592328,
  "HasAns_total": 45850,
  "NoAns_exact": 98.95512161699212,
  "NoAns_f1": 98.95512161699212,
  "NoAns_total": 23352,
  "best_exact": 90.77483309730933,
  "best_exact_thresh": 0.0,
  "best_f1": 94.94714391266305,
  "best_f1_thresh": 0.0
}

Comparison:

Model	f1 score
bert-base-spanish-wwm-cased-finetuned-spa-squad2-es	86.07
distill-bert-base-spanish-wwm-cased-finetuned-spa-squad2-es	94.94

So, yes, this version is even more accurate.

Model in action (in a Colab Notebook)

Set the context and ask some questions:

Run predictions:

Using Pipelines

Created by Manuel Romero/@mrm8488

Made with ♥ in Spain

3.6 KiB Raw Blame History