The current instance_score.summary is not readable #1588
Description
print("Instance Results:")
print(results.instance_scores.summary)
text_a text_a_type \
0 The drain is clogged with hair. It has to be cleaned. premise
1 Jane knocked on Susan's door but she did not answer. premise
2 Beth didn't get angry with Sally, who had cut her off, because she stopped and counted to ten. premise
3 No one joins Facebook to be sad and lonely. But a new study from the University of Wisconsin psychologist George Lincoln argues that that's exactly how it makes us feel. premise
4 The man couldn't lift his son because he was so heavy. premise
text_b text_b_type classes type_of_relation label \
0 The hair has to be cleaned. hypothesis [entailment, not entailment] entailment entailment
1 Susan did not answer. hypothesis [entailment, not entailment] entailment not entailment
2 Sally stopped and counted to ten. hypothesis [entailment, not entailment] entailment entailment
3 That's exactly how Facebook makes us feel. hypothesis [entailment, not entailment] entailment not entailment
4 The son was so heavy. hypothesis [entailment, not entailment] entailment not entailment
source \
0 [{'role': 'system', 'content': 'Given a premise and hypothesis cla
ssify the entailment of the hypothesis to one of entailment, not entailment.'}, {'role': 'user', 'content': 'premise: The drain is clogged with hair. It has to be cleaned.
hypothesis: The hair has to be cleaned.'}]
1 [{'role': 'system', 'content': 'Given a premise and hypothe
sis classify the entailment of the hypothesis to one of entailment, not entailment.'}, {'role': 'user', 'content': 'premise: Jane knocked on Susan's door but she did not answer.
hypothesis: Susan did not answer.'}]
2 [{'role': 'system', 'content': 'Given a premise and hypothesis classify the entailment of the hypothesis to one o
f entailment, not entailment.'}, {'role': 'user', 'content': 'premise: Beth didn't get angry with Sally, who had cut her off, because she stopped and counted to ten.
hypothesis: Sally stopped and counted to ten.'}]
3 [{'role': 'system', 'content': 'Given a premise and hypothesis classify the entailment of the hypothesis to one of entailment, not entailment.'}, {'role': 'user', 'content': 'premise: No one joins
Facebook to be sad and lonely. But a new study from the University of Wisconsin psychologist George Lincoln argues that that's exactly how it makes us feel.
hypothesis: That's exactly how Facebook makes us feel.'}]
4 [{'role': 'system', 'content': 'Given a premise and hypothesi
s classify the entailment of the hypothesis to one of entailment, not entailment.'}, {'role': 'user', 'content': 'premise: The man couldn't lift his son because he was so heavy.
hypothesis: The son was so heavy.'}]
metrics postprocessors target references \
0 [metrics.f1_micro, metrics.accuracy, metrics.f1_macro] [processors.take_first_non_empty_line, processors.lower_case_till_punc] entailment [entailment]
1 [metrics.f1_micro, metrics.accuracy, metrics.f1_macro] [processors.take_first_non_empty_line, processors.lower_case_till_punc] not entailment [not entailment]
2 [metrics.f1_micro, metrics.accuracy, metrics.f1_macro] [processors.take_first_non_empty_line, processors.lower_case_till_punc] entailment [entailment]
3 [metrics.f1_micro, metrics.accuracy, metrics.f1_macro] [processors.take_first_non_empty_line, processors.lower_case_till_punc] not entailment [not entailment]
4 [metrics.f1_micro, metrics.accuracy, metrics.f1_macro] [processors.take_first_non_empty_line, processors.lower_case_till_punc] not entailment [not entailment]
prediction \
0 **Given Premise:** The drain is clogged with hair. It has to be cleaned.\
n\n**Given Hypothesis:** The hair has to be cleaned.\n\n**Classification of the Entailment**: Both the premise and the hypothesis are saying the same thing - that there is a need to clean the hair. Th
erefore, we can say that the hypothesis is an **entailment** of the premise. The hypothesis is a consequence or a necessary condition of the premise, and it is not plausible to hold the premise withou
t the hypothesis being true.
1
According to the logic:\n\nSince the premise states that Jane knocked on Susan's door, and the hypothesis further states that Susan did
not answer, we can conclude that the hypothesis ENTAILS the premise.\n\nReasoning: If Susan had answered, the premise would not have been true. So, the lack of answer from Susan supports the hypothesi
s that Susan did not answer.
2 The hypothesis "Sally stopped and counted to ten" entails the premise "Beth didn't get angry with Sally, who had cut her off, because she stopped and counted to ten." It's a mathematical certainty
that Beth didn't get angry because Sally's actions were not a cause of her anger.\n\nHere's why the entailment works:\n\n- If Sally stopped and counted to ten, then any other action (like cutting off
Beth) would not cause her to get angry.\n- Therefore, Beth did not get angry with Sally because she stopped and counted to ten.\n\nThis is an example of a categorical syllogism: **All A, because B.**
(All that part G was part H)
3
**Premise:** No one joins Facebook to be sad and lonely.\n\n**Hypothesis:** That's exactly how Facebook makes us fee
l.\n\nThis is an example of an **entailment**. The premise logically guarantees the presence of a particular outcome (being sad and lonely) for Facebook usage, making it a significant predictor for Fa
cebook users' mental states.
4
The entailment is "not": The son was so heavy is an entailment of the premise. It contains the informa
tion that the son was heavy.
processed_prediction processed_references f1_macro f1_entailment score_name score accuracy f1_micro \
0 **given premise:** the drain is clogged with hair [entailment] 0 0 f1_micro 0 0 0
1 according to the logic: [not entailment] 0 NaN f1_micro 0 0 0
2 the hypothesis "sally stopped and counted to ten" entails the premise "beth didn't get angry with sally [entailment] 0 0 f1_micro 0 0 0
3 **premise:** no one joins facebook to be sad and lonely [not entailment] 0 NaN f1_micro 0 0 0
4 the entailment is "not": the son was so heavy is an entailment of the premise [not entailment] 0 NaN f1_micro 0 0 0
f1_not entailment
0 NaN
1 0
2 NaN
3 0
4 0
We should change it to be readable. E.g. show the source,references, prediction, processed_references,processed_predictions, score, score_name - maybe even in a line by line format because each of them is very long.