eleuther-pythia12b-hh-sft / sft-pythia-12b-5shot-shelloutput.txt
lomahony's picture
Upload 8 files
e12cdad
bootstrapping for stddev: perplexity
hf (pretrained=lomahony/eleuther-pythia12b-hh-sft,parallelize=True), limit: None, num_fewshot: 5, batch_size: 8
| Task |Version|Filter| Metric | Value | |Stderr|
|--------------|-------|------|---------------|------:|---|-----:|
|arc_challenge |Yaml |none |acc | 0.3712|± |0.0141|
| | |none |acc_norm | 0.4036|± |0.0143|
|arc_easy |Yaml |none |acc | 0.7201|± |0.0092|
| | |none |acc_norm | 0.7273|± |0.0091|
|boolq |Yaml |none |acc | 0.7220|± |0.0078|
|hellaswag |Yaml |none |acc | 0.5075|± |0.0050|
| | |none |acc_norm | 0.6749|± |0.0047|
|lambada_openai|Yaml |none |perplexity | 4.3599|± |0.1008|
| | |none |acc | 0.6697|± |0.0066|
|openbookqa |Yaml |none |acc | 0.3080|± |0.0207|
| | |none |acc_norm | 0.4040|± |0.0220|
|piqa |Yaml |none |acc | 0.7677|± |0.0099|
| | |none |acc_norm | 0.7704|± |0.0098|
|sciq |Yaml |none |acc | 0.9450|± |0.0072|
| | |none |acc_norm | 0.9480|± |0.0070|
|wikitext |Yaml |none |word_perplexity|16.5611| | |
| | |none |byte_perplexity| 1.5884| | |
| | |none |bits_per_byte | 0.6676| | |
|winogrande |Yaml |none |acc | 0.6346|± |0.0135|