Views
No views yet
batch_size = 96
n_epochs = 2
base_LM_model = "roberta-base"
max_seq_len = 386
learning_rate = 3e-5
lr_schedule = LinearWarmup
warmup_proportion = 0.2
doc_stride=128
max_query_length=641reader = FARMReader(model_name_or_path="deepset/roberta-base-squad2")
2# or
3reader = TransformersReader(model_name_or_path="deepset/roberta-base-squad2",tokenizer="deepset/roberta-base-squad2")roberta-base-squad2 being used for Question Answering, check out the Tutorials in Haystack Documentation1from transformers import AutoModelForQuestionAnswering, AutoTokenizer, pipeline
2
3model_name = "deepset/roberta-base-squad2"
4
5# a) Get predictions
6nlp = pipeline('question-answering', model=model_name, tokenizer=model_name)
7QA_input = {
8 'question': 'Why is model conversion important?',
9 'context': 'The option to convert models between FARM and transformers gives freedom to the user and let people easily switch between frameworks.'
10}
11res = nlp(QA_input)
12
13# b) Load model & tokenizer
14model = AutoModelForQuestionAnswering.from_pretrained(model_name)
15tokenizer = AutoTokenizer.from_pretrained(model_name)"exact": 79.87029394424324,
"f1": 82.91251169582613,
"total": 11873,
"HasAns_exact": 77.93522267206478,
"HasAns_f1": 84.02838248389763,
"HasAns_total": 5928,
"NoAns_exact": 81.79983179142137,
"NoAns_f1": 81.79983179142137,
"NoAns_total": 5945