eleuther-pythia6.9b-hh-dpo / base-6.9b-eval-files /EleutherAI-pythia-6.9b-5shot-shelloutput.txt
lomahony's picture
Upload 8 files
f9eae2b
bootstrapping for stddev: perplexity
hf (pretrained=EleutherAI/pythia-6.9b), limit: None, num_fewshot: 5, batch_size: 4
| Task |Version|Filter| Metric | Value | |Stderr|
|--------------|-------|------|---------------|------:|---|-----:|
|arc_challenge |Yaml |none |acc | 0.3541|± |0.0140|
| | |none |acc_norm | 0.3891|± |0.0142|
|arc_easy |Yaml |none |acc | 0.6944|± |0.0095|
| | |none |acc_norm | 0.7045|± |0.0094|
|boolq |Yaml |none |acc | 0.6575|± |0.0083|
|hellaswag |Yaml |none |acc | 0.4804|± |0.0050|
| | |none |acc_norm | 0.6509|± |0.0048|
|lambada_openai|Yaml |none |perplexity | 5.6328|± |0.1331|
| | |none |acc | 0.6231|± |0.0068|
|openbookqa |Yaml |none |acc | 0.2800|± |0.0201|
| | |none |acc_norm | 0.3540|± |0.0214|
|piqa |Yaml |none |acc | 0.7644|± |0.0099|
| | |none |acc_norm | 0.7633|± |0.0099|
|sciq |Yaml |none |acc | 0.9470|± |0.0071|
| | |none |acc_norm | 0.9550|± |0.0066|
|wikitext |Yaml |none |word_perplexity|17.6830| | |
| | |none |byte_perplexity| 1.6057| | |
| | |none |bits_per_byte | 0.6832| | |
|winogrande |Yaml |none |acc | 0.6069|± |0.0137|