Hugging Face's logo

datablations
/
lm1-misc-pile

TensorBoard
Model card Files Metrics Community
Muennighoff commited on
Commit
956e279
•
1 Parent(s): 0461a88

Add

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_0.json +1 -0
  2. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_1.json +1 -0
  3. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_2.json +1 -0
  4. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_3.json +1 -0
  5. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_4.json +1 -0
  6. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_5.json +1 -0
  7. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_0.json +1 -0
  8. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_1.json +1 -0
  9. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_2.json +1 -0
  10. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_3.json +1 -0
  11. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_4.json +1 -0
  12. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_5.json +1 -0
  13. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_0.json +1 -0
  14. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_1.json +1 -0
  15. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_2.json +1 -0
  16. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_3.json +1 -0
  17. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_4.json +1 -0
  18. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_5.json +1 -0
  19. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_0.json +1 -0
  20. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_1.json +1 -0
  21. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_2.json +1 -0
  22. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_3.json +1 -0
  23. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_4.json +1 -0
  24. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_5.json +1 -0
  25. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_0.json +1 -0
  26. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_1.json +1 -0
  27. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_2.json +1 -0
  28. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_3.json +1 -0
  29. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_4.json +1 -0
  30. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_5.json +1 -0
  31. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_0.json +1 -0
  32. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_1.json +1 -0
  33. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_2.json +1 -0
  34. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_3.json +1 -0
  35. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_4.json +1 -0
  36. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_5.json +1 -0
  37. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_0.json +1 -0
  38. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_1.json +1 -0
  39. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_2.json +1 -0
  40. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_3.json +1 -0
  41. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_4.json +1 -0
  42. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_5.json +1 -0
  43. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_0.json +1 -0
  44. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_1.json +1 -0
  45. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_2.json +1 -0
  46. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_3.json +1 -0
  47. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_4.json +1 -0
  48. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_5.json +1 -0
  49. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_tldr_en_0.json +1 -0
  50. 2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_tldr_en_1.json +1 -0
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_0.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "bleu": 0.2631325271644541, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.015483068626320248}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_precision": 0.07301531048529383, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.00274558662513083}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_recall": 0.25682902081294867, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004665164573472525}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_fmeasure": 0.09460580374950654, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.002070112611762097}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_precision": 0.03135773359810665, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0015196010938366357}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_recall": 0.1190793841560146, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.00314719277960027}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_fmeasure": 0.042171217148328305, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0012014862715668082}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_precision": 0.06995138221881579, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.002585855855883416}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_recall": 0.25034386254696944, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004558097215891303}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_fmeasure": 0.09114864432806526, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0019292823881781717}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_precision": 0.06923355554637999, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0026028539500661096}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_recall": 0.24434142921166224, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004406903022665613}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_fmeasure": 0.08971837624266346, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0019451835942869572}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 0, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_1.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "bleu": 0.6319598443147796, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.054301599489699426}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_precision": 0.1449802647542582, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.004835369406099664}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_recall": 0.3058074823164192, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.005348054701943042}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_fmeasure": 0.15671464020064418, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0037866455873768667}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_precision": 0.07188064353062813, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.003205240820105057}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_recall": 0.15268344072314297, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0036666323434365483}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_fmeasure": 0.07720455939389521, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0024879092576665586}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_precision": 0.13053458022078224, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.004368574859372304}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_recall": 0.2854132484030163, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004943844091747833}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_fmeasure": 0.1416205080340026, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003272829669505245}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_precision": 0.13328996266673795, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.00447012253890352}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_recall": 0.28672965869175415, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004941088950081213}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_fmeasure": 0.14390649445247156, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0033614402461819514}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 1, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_2.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "bleu": 0.6929720546915326, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.03574586631609437}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_precision": 0.1667489031816748, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0053129091662153394}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_recall": 0.312906549882645, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.005184017730529131}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_fmeasure": 0.17119248724834354, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004079889456936537}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_precision": 0.08932829811722608, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.003614965400443546}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_recall": 0.16318957259394432, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.003657806537105316}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_fmeasure": 0.0892712295730392, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.002768450465820049}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_precision": 0.14872990634724728, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0047109702306293065}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_recall": 0.2907739402970601, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004717972403865755}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_fmeasure": 0.15389013740086638, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0034924548969777276}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_precision": 0.15270363372360954, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.004857391805953228}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_recall": 0.29443587522831705, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0047910204345461265}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_fmeasure": 0.15719053925266066, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0036077850902892132}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 2, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_3.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "bleu": 0.7793413569965535, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.04401095166360687}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_precision": 0.1851553824117999, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.005739543098874004}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_recall": 0.32096672875237564, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0052205548821339254}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_fmeasure": 0.18098791905747405, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004171835434472728}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_precision": 0.0987958700237029, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0039690050533069485}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_recall": 0.1682300557957493, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.003645629757025838}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_fmeasure": 0.09448880082729369, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0028077781148065834}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_precision": 0.16591920090104267, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005175506248042414}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_recall": 0.2968310250157565, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004725975660016804}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_fmeasure": 0.1624665012419433, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003603720803604403}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_precision": 0.17017226689910603, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005310411726141506}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_recall": 0.3009925111403994, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004808313258795627}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_fmeasure": 0.16603488625521012, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.00370981876565165}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 3, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_4.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "bleu": 0.9977078657471178, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.07887847921625667}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_precision": 0.2059293837362699, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.005930727757990343}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_recall": 0.34195039139075256, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.00530284692088583}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_fmeasure": 0.198981310213424, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004371062695903911}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_precision": 0.1132664784056494, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.004195666504864259}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_recall": 0.1817959980117022, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0038979348851944664}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_fmeasure": 0.10662610861919704, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0030973990146846123}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_precision": 0.18268156276094377, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005299944414313383}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_recall": 0.31336499591459327, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004779934135701508}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_fmeasure": 0.17685938846601176, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003750212526612896}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_precision": 0.1894427258620293, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005520444753900842}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_recall": 0.318922687382484, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004854844424024468}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_fmeasure": 0.18213026739107405, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.003903040459492487}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 4, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_PALM_prompt_5.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "bleu": 1.0329034246178166, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.06101593441964185}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_precision": 0.21222641779096513, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.00600931540723154}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_recall": 0.34835245504750145, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.005362537610194471}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge1_fmeasure": 0.20798260134057678, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004573750265687856}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_precision": 0.1146743488998646, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.004087815577545942}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_recall": 0.18829629299952763, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.003958867454191072}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rouge2_fmeasure": 0.11155001600158532, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.003149804901867283}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_precision": 0.18655761229031262, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.00531944527507642}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_recall": 0.31868310470746836, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0048359994015566}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeL_fmeasure": 0.18365484301367235, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003888537533617311}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_precision": 0.19370708600762668, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005530833290674727}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_recall": 0.32480224216024617, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004914527796380582}, {"task_name": "GEM/web_nlg_en", "prompt_name": "PALM_prompt", "rougeLsum_fmeasure": 0.18949373871891598, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "3e41305c-5461-4cf3-853d-8a6fb5747623", "prompt_jinja": "I will verbalize an abstract representation of a sentence in natural language. To do so, I will first show the representation and then the natural language. The text needs to include all of the information in the representation.\n\n{{input | join(\", \")}} {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0040492718760054645}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 5, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_0.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_precision": 0.04755573080486142, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0014943001894154296}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_recall": 0.17639887782882707, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.002781073860145518}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_fmeasure": 0.06011830321976551, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0012407590197264515}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_precision": 0.004883934331922545, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.00048234829631821883}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_recall": 0.021186182792688506, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.001434173564522441}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_fmeasure": 0.006467662095401446, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.000489691287606407}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_precision": 0.04189513804854788, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0012678933112630836}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_recall": 0.16267055979911485, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0025808862203828926}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_fmeasure": 0.053604807011045266, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0010394560540794644}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_precision": 0.042939863244687196, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0014045908608215807}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_recall": 0.15313515752116336, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0024831027228757337}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_fmeasure": 0.0533718368033741, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.001156691298097272}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "bleu": 0.09248512201038334, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.007365914003567337}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 0, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_1.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_precision": 0.5262824609719964, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.006444296448467724}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_recall": 0.3821460760323023, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004509002868672824}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_fmeasure": 0.3931164713482728, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004449720706795851}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_precision": 0.27540644166441025, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005320580176652835}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_recall": 0.19116976243278833, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.003730111195098187}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_fmeasure": 0.19879100920250162, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0037765603841301582}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_precision": 0.4351652371468361, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005820796414201653}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_recall": 0.318491733812937, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004091119706038044}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_fmeasure": 0.3236587681409084, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003971936119705106}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_precision": 0.46057375818904694, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.006023827761868843}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_recall": 0.3319940090961664, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0040764844035161025}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_fmeasure": 0.3408093729612861, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0040385638047001175}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "bleu": 6.712294977029452, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.1565455932084625}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 1, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_2.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_precision": 0.6705357937808699, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.005358560021307114}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_recall": 0.46180595409675135, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004723115930056159}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_fmeasure": 0.5043240560808788, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004225044305963183}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_precision": 0.39660323883414605, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005295237552613934}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_recall": 0.26689626240946507, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0042439563745714755}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_fmeasure": 0.2915762298992619, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004140723345919459}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_precision": 0.561666433050668, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005237299955479755}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_recall": 0.38734294263241686, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004427321763654563}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_fmeasure": 0.4212431053828867, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.004078947666860322}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_precision": 0.5917250537465801, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005306074623393257}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_recall": 0.4035455233772213, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004393553894619333}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_fmeasure": 0.4410013874398624, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.004006777615967865}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "bleu": 12.094423896118393, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.25654719532533016}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 2, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_3.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_precision": 0.6808220065103112, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.005316453481045833}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_recall": 0.4663807888071689, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.00479470778423173}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_fmeasure": 0.5116659645680365, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004236807407419154}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_precision": 0.41361447320688666, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.00536282608667289}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_recall": 0.2769120315587748, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004340204678489145}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_fmeasure": 0.30407197816456755, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004229553051713374}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_precision": 0.5780584368527751, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005253922993906422}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_recall": 0.39570598608147833, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004515550109669547}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_fmeasure": 0.4326984289599196, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.00411134294399591}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_precision": 0.6039999493756314, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005331495580044351}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_recall": 0.40918584407900555, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004456419514851121}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_fmeasure": 0.4493663995499122, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.00403615125088154}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "bleu": 13.050064301025847, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.3649807787867134}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 3, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_4.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_precision": 0.700573219004743, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0049679172668200475}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_recall": 0.46073001653993867, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004829895393678769}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_fmeasure": 0.517783062988282, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.00413971275530478}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_precision": 0.4261607573249779, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005214349556938651}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_recall": 0.2763776678769553, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004355931558733263}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_fmeasure": 0.30928733049941687, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004165946092527738}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_precision": 0.5941745918368991, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.00501648243119666}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_recall": 0.39017928990272394, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004514712832266304}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_fmeasure": 0.43726162609761876, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.004024981693661591}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_precision": 0.6209033973394869, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005036518469962893}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_recall": 0.4040127169874031, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004497792578185556}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_fmeasure": 0.45449128765435426, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0039619267882531365}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "bleu": 12.143660238953467, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.2618377981953508}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 4, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_explicit-graph-description2_5.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_precision": 0.6990114866921079, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.004888887009805971}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_recall": 0.4562737779493208, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004809549420929517}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge1_fmeasure": 0.5150940910808839, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004054282953161343}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_precision": 0.4307024571566198, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005276581160539604}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_recall": 0.27484552178475224, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004350406168331502}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rouge2_fmeasure": 0.30983810327293143, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004135319155984754}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_precision": 0.5974681521724369, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0050614329568243405}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_recall": 0.3878291391697164, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0045241666773479135}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeL_fmeasure": 0.4374551467308437, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003995930860580631}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_precision": 0.623588192584714, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005026770729643474}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_recall": 0.402310523858126, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004481446026309217}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "rougeLsum_fmeasure": 0.4548550330143833, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.003899518686789911}, {"task_name": "GEM/web_nlg_en", "prompt_name": "explicit-graph-description2", "bleu": 12.187686476415259, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "afeec167-f75f-4687-a775-1efde7d04780", "prompt_jinja": "{{input | join(\", \")}}. \n\nThe above is a set of subject | predicate | object expressions separated by commas: \nWrite all the information in proper sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.357825457832992}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 5, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_0.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "bleu": 0.22078716311166532, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.03101252566016281}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_precision": 0.04336717623202954, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0011021542631244458}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_recall": 0.2688220997120123, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004486164427127026}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_fmeasure": 0.06826095797871463, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0014005068999256936}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_precision": 0.008392895002023323, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0004718780469155826}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_recall": 0.06093035970117915, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0028298512424470594}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_fmeasure": 0.013978494461989985, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.000730999460164706}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_precision": 0.039924133583051816, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0009183561667963128}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_recall": 0.25509796128052686, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004078923005175043}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_fmeasure": 0.06352298772020437, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0011852080942429933}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_precision": 0.03405470850263721, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0010040982903321283}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_recall": 0.20903086908834975, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004036260934244987}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_fmeasure": 0.052807846723970854, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0012508311968299822}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 0, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_1.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "bleu": 6.2064513888831625, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.24556514108315913}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_precision": 0.4759462678430596, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.006340986072606315}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_recall": 0.3868799134851979, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004804061068719801}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_fmeasure": 0.3735412029791271, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004526482791493285}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_precision": 0.24338523022257152, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005021543487605631}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_recall": 0.19210997327178064, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0037951744740514345}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_fmeasure": 0.1854092786101686, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.003662828553417364}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_precision": 0.395042471506904, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005657862769855998}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_recall": 0.32359808853904054, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004242717659431168}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_fmeasure": 0.30853968812248894, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003941694225820738}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_precision": 0.41925715302149796, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0058792012178161}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_recall": 0.33858822427291213, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004296499170924593}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_fmeasure": 0.3260209932351036, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.004068827011546382}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 1, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_2.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "bleu": 9.029832541252915, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.5906741044962391}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_precision": 0.5881379118156979, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.00607138355041428}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_recall": 0.46525909590251635, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004645617067304566}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_fmeasure": 0.4660447970734647, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004437680629867758}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_precision": 0.33879491010289237, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005284190061409969}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_recall": 0.26141501301429376, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004135663693546733}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_fmeasure": 0.2616858898937414, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0040067556798890165}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_precision": 0.49304024171616717, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0056223879832542605}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_recall": 0.39132153508338124, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0043344385768999145}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_fmeasure": 0.3888379311609116, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.004095298102298575}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_precision": 0.5200784980417149, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005796072025741954}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_recall": 0.40779591077095406, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004326197563353865}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_fmeasure": 0.40805303250091984, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.004125467023459015}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 2, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_3.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "bleu": 11.462789049772756, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.544286138229245}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_precision": 0.6131002474315037, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0058625863987195525}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_recall": 0.4740608580782679, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004745441576492749}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_fmeasure": 0.4864605359125179, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0044290307128628205}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_precision": 0.3616792010621346, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0053466472768704055}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_recall": 0.27449730351862534, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004320167990529969}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_fmeasure": 0.28201013635702454, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004233353412300328}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_precision": 0.5175613257182873, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005543135489613363}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_recall": 0.4000231071212494, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004426986087663464}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_fmeasure": 0.40888811724077884, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.004166479643581242}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_precision": 0.5442933154452568, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005652343424098652}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_recall": 0.41733108185520645, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0044137293630639955}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_fmeasure": 0.4285410880798501, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.004174530523840161}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 3, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_4.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "bleu": 13.195903034981136, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.27680487078940497}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_precision": 0.6266260417490315, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.005428501367170535}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_recall": 0.47379894820307406, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004766471434240934}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_fmeasure": 0.4954977191144857, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004170214447974351}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_precision": 0.37127027022389625, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005108897459139562}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_recall": 0.2768160149581001, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004224704962527318}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_fmeasure": 0.288273235913836, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.003995368654583399}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_precision": 0.5289133337530646, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005206394449338708}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_recall": 0.40011570414743575, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0044079592358739475}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_fmeasure": 0.41670064949801117, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0039385953564280325}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_precision": 0.5579736430717384, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005280103100263265}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_recall": 0.4179503942991226, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004383763303015715}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_fmeasure": 0.4375131145007485, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0039093732406277545}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 4, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_implicit-graph-description_5.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "bleu": 13.619640181036603, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.3398573433105108}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_precision": 0.6435523638173357, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.005191365215959936}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_recall": 0.4746336745456459, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004861740789134235}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge1_fmeasure": 0.5054664942024651, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004096959648978707}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_precision": 0.3862702704182125, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005096309491585679}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_recall": 0.28171369115294387, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004359192359656378}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rouge2_fmeasure": 0.2977913696390576, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004026220746231584}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_precision": 0.5457148890002358, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005130660710283408}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_recall": 0.40179512645802773, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004520144367899868}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeL_fmeasure": 0.4262782517140752, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003940419226504188}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_precision": 0.5740054014119437, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005111288294537433}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_recall": 0.42054413160499, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004531431662740779}, {"task_name": "GEM/web_nlg_en", "prompt_name": "implicit-graph-description", "rougeLsum_fmeasure": 0.44760311861187624, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "38342608-5cd7-4ce7-b2e1-905ecd7f4c80", "prompt_jinja": "{{input | join(\"; \")}}\nThe above is a collection of relations. Write descriptive English that contains this information.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.003894052257554968}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 5, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_0.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_precision": 0.09235409599713389, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.002123164893741087}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_recall": 0.39991815334946373, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.005500287721993977}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_fmeasure": 0.13444220603715523, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.002296851249870823}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_precision": 0.028916925815618007, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0012272881918049938}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_recall": 0.14475568918624443, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.003451599107114426}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_fmeasure": 0.0435885480008138, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0013176989907165808}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_precision": 0.07687375395422678, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0017296683313182154}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_recall": 0.3438951712475559, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004836102226003126}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_fmeasure": 0.1118681637273035, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.001820062964373532}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_precision": 0.08000935297969644, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0018429012610192957}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_recall": 0.3458047115902232, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004678038749731419}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_fmeasure": 0.1158716704132779, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0019872759622980404}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "bleu": 0.2286722670848296, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.031899806884053854}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 0, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_1.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_precision": 0.4857482679204505, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.005969273957403387}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_recall": 0.47479496265753507, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.005028089432620777}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_fmeasure": 0.42059990211741793, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004448317895973875}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_precision": 0.2515271091606676, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.004629940725817203}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_recall": 0.246657962427562, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004100703509821082}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_fmeasure": 0.2150475903823648, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.003652352532292937}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_precision": 0.3984440027781437, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005303560272649928}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_recall": 0.3912215165779194, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004473892199929174}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_fmeasure": 0.3430358661846022, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0038919150224397364}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_precision": 0.42359884183510665, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.00548458889679864}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_recall": 0.41436689313466735, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004593672851400749}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_fmeasure": 0.3641538291748135, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.003964447780428173}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "bleu": 5.906545551936352, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.20043699662445755}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 1, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_2.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_precision": 0.6030219591627299, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.005923870140276083}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_recall": 0.498751333697349, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004873991369817728}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_fmeasure": 0.49509937902003354, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004539589245202305}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_precision": 0.34173934445628185, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005085688962735734}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_recall": 0.2816699812579884, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004347883852138146}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_fmeasure": 0.27803595379552354, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004147494942597356}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_precision": 0.49299813204584697, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005378320646499662}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_recall": 0.41097925347656994, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004485206918199883}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_fmeasure": 0.4045496015202703, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.004164809990370727}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_precision": 0.523044369743198, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005549967316924623}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_recall": 0.43162764603022075, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004500211133471427}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_fmeasure": 0.42699481188468885, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.004184692525573101}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "bleu": 9.305449250220704, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.6323942408861912}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 2, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_3.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_precision": 0.6214005744975052, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.00558395872620154}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_recall": 0.494594254641018, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004845836889627149}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_fmeasure": 0.5020487201575455, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0042836658596996134}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_precision": 0.3590485394733236, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005090410204360709}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_recall": 0.28301099553882336, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.00425896353194711}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_fmeasure": 0.28599053859644635, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004030001639965491}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_precision": 0.510942374391943, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005195057767003233}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_recall": 0.40900913172027126, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004432745886897139}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_fmeasure": 0.4122334559245298, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003959877746335585}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_precision": 0.541991926404449, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005351986316165339}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_recall": 0.4290124848300319, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004439413695000747}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_fmeasure": 0.43517942057846215, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.00397942507447921}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "bleu": 11.232752470980182, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.6070465163626537}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 3, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_4.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_precision": 0.6408314541854252, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.00548394048208633}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_recall": 0.4859786319072553, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004814120211184008}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_fmeasure": 0.5082034100540427, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004271012267039298}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_precision": 0.3714441762057441, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005028548667827809}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_recall": 0.2783546549864995, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004228515385743316}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_fmeasure": 0.290196814250039, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004005574732772928}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_precision": 0.5254254922724889, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005210124779354852}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_recall": 0.39928885031845557, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004413658106389613}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_fmeasure": 0.41526601467877283, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0039890190193175926}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_precision": 0.5568560862687746, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005295897424171165}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_recall": 0.4196394062285843, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004404374764675378}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_fmeasure": 0.43851206335898485, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.003962170479973872}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "bleu": 12.695732889726765, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.648263357957855}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 4, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_non-explicit-description_5.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_precision": 0.6487959831097037, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.005399289731985388}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_recall": 0.48623363033328076, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004778750202594707}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge1_fmeasure": 0.5117493091488112, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004165942578056248}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_precision": 0.3776843714696378, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005058278200508039}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_recall": 0.2801932440661686, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0042015972775485575}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rouge2_fmeasure": 0.2931459742422919, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.003948968688944467}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_precision": 0.5318684015270567, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005121981416572071}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_recall": 0.4000892803874322, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004383599495183284}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeL_fmeasure": 0.41851527317921877, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.003913262097008211}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_precision": 0.5643403370955602, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005237099171215742}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_recall": 0.4197849739822551, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.00438949794350207}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "rougeLsum_fmeasure": 0.4418102169986528, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.003887615682911017}, {"task_name": "GEM/web_nlg_en", "prompt_name": "non-explicit-description", "bleu": 13.159654141205856, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "9415bd8a-685f-4fa4-803a-f09bd47d4603", "prompt_jinja": "I am taking this tabular data, where each row is separated by a \"|\" and expresses a relation between an object and a predicate : {{input | join(\", \")}}. \n\nNow, I will produce a description of the tabular data using English sentences. {% for i in references %}\n ||| {{ i }} \n{% endfor %}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.3948757014086876}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 5, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_0.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_precision": 0.06447087283897882, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0014590914914983882}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_recall": 0.4558011383728955, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0071156347222442}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_fmeasure": 0.10810806384928827, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0022120236049110935}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_precision": 0.01947895038358896, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0006829809352677969}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_recall": 0.1473034605928463, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.003749142485679465}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_fmeasure": 0.032669341341239076, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0010490179341564594}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_precision": 0.048359420119038043, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0009207227437773181}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_recall": 0.3725994892406033, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.005880967307305537}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_fmeasure": 0.08203945806881063, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.001416260235673195}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_precision": 0.05597766416506184, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.001277509974018661}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_recall": 0.40168336120338893, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.00637370623271962}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_fmeasure": 0.09395028450927674, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0019379616517689725}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "bleu": 0.2838534319721657, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.024626840811213652}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 0, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_1.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_precision": 0.515101106223276, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0061065748509348304}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_recall": 0.4925172064677324, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004811947371245265}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_fmeasure": 0.4484759030001043, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0046496418757624655}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_precision": 0.27412311564080366, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.004771745214030575}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_recall": 0.26074415987734006, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004045350822111356}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_fmeasure": 0.23529342485436736, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0038015211409521948}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_precision": 0.41984814704853923, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005419852908600043}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_recall": 0.40692973250413594, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004381139705596441}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_fmeasure": 0.3649455118211661, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0040716056405398875}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_precision": 0.444447322893853, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005598457656251664}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_recall": 0.42460953739901913, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004369920161304386}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_fmeasure": 0.38484826255511656, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.004176128500327213}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "bleu": 5.604651790563788, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.33170459892685983}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 1, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_2.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_precision": 0.5988458774735271, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.006065429124056104}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_recall": 0.5051251938340441, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004824474720978224}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_fmeasure": 0.4943668950624586, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004580813756634929}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_precision": 0.33919450998994116, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005172819110529959}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_recall": 0.28283630209757277, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.00429437705388263}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_fmeasure": 0.27605470258504766, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004142652419009752}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_precision": 0.48861507185343345, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005548502846526632}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_recall": 0.41625061047105966, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004486998905477303}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_fmeasure": 0.40264507869420074, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.00419340631440858}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_precision": 0.519253201762426, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.00568847145833182}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_recall": 0.435844877359072, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004476472188770956}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_fmeasure": 0.4254164205944382, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.004212261673169034}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "bleu": 8.615673013549769, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.447440485410148}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 2, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_3.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_precision": 0.6405218569752045, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.00571091746612946}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_recall": 0.48478865759540896, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004869101818120115}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_fmeasure": 0.5043883739223508, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004458819212755122}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_precision": 0.3686287974389768, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005188823520619212}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_recall": 0.276378642516858, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004268288307454759}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_fmeasure": 0.2863559004655948, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004132657814507632}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_precision": 0.5240261149211461, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005319993158590117}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_recall": 0.39948134144087455, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0044547869019207305}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_fmeasure": 0.4120533408244971, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.004106047847002439}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_precision": 0.5557134083849269, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0055023282986634454}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_recall": 0.4172210306339939, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004450975951856631}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_fmeasure": 0.43388866185707814, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.004135112955801659}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "bleu": 11.110519188887649, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.52295909337497}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 3, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_4.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_precision": 0.66995722254869, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0052094650430898086}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_recall": 0.47914137998448814, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0047870042792149}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_fmeasure": 0.5180969516792059, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.004189427957426417}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_precision": 0.3915009591541496, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.00507371115460548}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_recall": 0.2768737640978803, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.00430470760346202}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_fmeasure": 0.29795300756428195, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004099730361476593}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_precision": 0.5458856984805986, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005028084082936345}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_recall": 0.39221537131903234, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004468924978865893}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_fmeasure": 0.42128033557057226, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.004002651728774077}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_precision": 0.5794561378932997, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005103634599096425}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_recall": 0.41210371956869646, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004453761428101923}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_fmeasure": 0.44485211690432075, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.003945111488311744}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "bleu": 12.519280690010929, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.3333969252082408}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 4, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-web_nlg_en_very-explicit-description_5.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_precision": 0.682373923286837, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.00516060397035127}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_recall": 0.4713995404298349, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.004906886331479491}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge1_fmeasure": 0.5159798396890325, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0042378651081273816}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_precision": 0.40406716412567933, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.005216340252635802}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_recall": 0.275345792931267, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.004393892177008509}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rouge2_fmeasure": 0.300194990426099, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.004145349477312573}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_precision": 0.5630405519088766, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.005102043855168231}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_recall": 0.38936478291695387, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.004557373177394848}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeL_fmeasure": 0.4246354304238641, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.004087168248306567}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_precision": 0.5960875125629144, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.005140712888518856}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_recall": 0.4078049819358144, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.004520115122872941}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "rougeLsum_fmeasure": 0.4467461307083424, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.003997888286874406}, {"task_name": "GEM/web_nlg_en", "prompt_name": "very-explicit-description", "bleu": 12.690188910837392, "fixed_answer_choice_list": null, "dataset_path": "GEM/web_nlg", "dataset_name": "en", "subset": null, "prompt_id": "426b682e-e801-4e8d-9ac3-5b676c9d3da2", "prompt_jinja": "A semantic triple is the atomic data entity in the Resource Description Framework (RDF) data model. As its name indicates, a triple is a set of three entities that codifies a statement about semantic data in the form of subject\u2013predicate\u2013object expressions. (e.g., \"Bob | is | 35\", or \"Bob | knows | John\"). \n\nA graph can be formed from a set of these triples. An example is {{input | join(\", \")}}. \n\nWrite grammatical text expressing all the relations succinctly and fluently.\n{% for i in references %}\n ||| {{ i }} \n{% endfor %}\n\n", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.28842273502317445}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 5, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_0.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_precision": 0.17944982620151825, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0019738049861084787}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_recall": 0.31392424087953436, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0028010306395309355}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_fmeasure": 0.21229608722254698, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0018911386726812303}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_precision": 0.03993898259187477, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0008518771365716035}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_recall": 0.07260259108150281, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0016432488281853878}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_fmeasure": 0.047439651785124645, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0009516875666742553}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_precision": 0.12572341758536942, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0012754167786637968}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_recall": 0.22851195511931727, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0021669353877673005}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_fmeasure": 0.15044112402271614, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0012475981593567816}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_precision": 0.16584885644997316, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0018114283517013234}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_recall": 0.2914092986653983, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.002619320469773644}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_fmeasure": 0.19643089116407775, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.001736720127476052}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "bleu": 2.078863824551135, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.08330163365121163}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 0, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_1.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_precision": 0.166305428507607, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0022062098631351968}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_recall": 0.2591678442905207, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0028472370580069056}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_fmeasure": 0.18339675598814223, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0019627555222683}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_precision": 0.032852922070060984, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0009454307433321334}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_recall": 0.05283614751346448, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.001433877665093853}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_fmeasure": 0.03614126849271882, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0009064689516384773}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_precision": 0.12021311578635133, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.001524545131273604}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_recall": 0.19278584644274216, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0021080039019195974}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_fmeasure": 0.13335554603361374, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0013046120546550155}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_precision": 0.15399841304150944, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.002032676997636144}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_recall": 0.24146920883940856, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.002662918829815805}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_fmeasure": 0.17017346113673246, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.00180595532677197}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "bleu": 1.72948259975478, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.051005040267799555}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 1, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_2.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_precision": 0.17529080817626147, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.00253379174585785}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_recall": 0.24654151460153587, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0028039306884267124}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_fmeasure": 0.18122204614486226, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0019713166292355793}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_precision": 0.036236657014771786, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0011758048358293301}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_recall": 0.05000815873043061, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.001397774717370768}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_fmeasure": 0.036087345347192205, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0009339729811344016}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_precision": 0.13014649831341282, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0018830365813501883}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_recall": 0.18642267598578363, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0020866432169766927}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_fmeasure": 0.13448544925044303, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.001342381669303193}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_precision": 0.16252327563684557, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.002364224902859441}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_recall": 0.22906152256564155, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0026087771740418804}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_fmeasure": 0.16796642541404475, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0018231044266279908}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "bleu": 1.7994940058131819, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.06735755594613854}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 2, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_3.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_precision": 0.1603998208563856, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.002932268910383442}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_recall": 0.20047482794881774, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0030764914583320175}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_fmeasure": 0.15199386706163123, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.002184056903266182}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_precision": 0.03537422537231008, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.001430374494319823}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_recall": 0.042072255613979134, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0013776186679184306}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_fmeasure": 0.031214635716306602, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0009373230961999606}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_precision": 0.12314016388275628, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0023464099851487584}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_recall": 0.15400469430404082, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0023450819349651194}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_fmeasure": 0.11530490214541701, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0015879752337995007}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_precision": 0.14859177384694494, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.002743279966132276}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_recall": 0.18612606421362016, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0028620644114185564}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_fmeasure": 0.1407243470689772, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0020201191591159716}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "bleu": 1.8094445237099583, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.06546698898519207}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 3, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_4.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_precision": 0.05917236382499021, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.002413272941068204}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_recall": 0.06666214305118287, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.00255340047793962}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_fmeasure": 0.051318953568170636, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0018669115061265386}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_precision": 0.013424205973822007, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0009751838857259187}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_recall": 0.01593931679588193, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0010273118038364056}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_fmeasure": 0.011638473351550615, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0006994551499063337}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_precision": 0.04644867083109973, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.001960516070504733}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_recall": 0.05143460229100747, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0019673266881076153}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_fmeasure": 0.039397782933602514, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0014126146018284696}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_precision": 0.05482673636274537, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.002244807970809181}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_recall": 0.06175321584900349, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0023715490442946553}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_fmeasure": 0.047468639548701494, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.001723259081997461}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "bleu": 0.1901093768303124, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.0182638181824565}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 4, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_article_summary_en_5.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_precision": 0.012716826474184926, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0014829658601519474}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_recall": 0.010754490521196739, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0011721536389422765}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge1_fmeasure": 0.008420330303836826, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0008572199651284122}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_precision": 0.00269710055327306, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0004665933402628681}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_recall": 0.0030878018550372093, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0005452140122555941}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rouge2_fmeasure": 0.0021281402364193283, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0003007767206985946}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_precision": 0.010312677790003952, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0012931686824202392}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_recall": 0.00828144194714077, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0009177036946589218}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeL_fmeasure": 0.006420199943171878, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0006542725812186021}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_precision": 0.011874076078290611, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0014222369864070887}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_recall": 0.009838223210952733, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0010793291142392508}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "rougeLsum_fmeasure": 0.007687814233842553, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0007856808016922665}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "article_summary_en", "bleu": 1.2581858244931555e-10, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "2038df7b-5420-4a33-87ec-09715419deef", "prompt_jinja": "Article in English: {{source}}\n\nSummary in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 1.5262857510956952e-09}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 5, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_0.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_precision": 0.10866990587172463, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0018725879698033455}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_recall": 0.1455391440670949, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0023576064226689714}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_fmeasure": 0.11440813592687969, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.001790995074200021}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_precision": 0.01652610652264889, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0006419029635214541}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_recall": 0.023584402514889233, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0009957317422302626}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_fmeasure": 0.017822732715514513, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0006743075533390354}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_precision": 0.09179627740172516, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.001467566887892716}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_recall": 0.12486912237053059, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0019541428043311653}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_fmeasure": 0.097076820121822, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0014067624557563508}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_precision": 0.10107941143612967, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0017331892649792644}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_recall": 0.13596469032060313, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.00219624032655005}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_fmeasure": 0.10659223653356883, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.001658458438378257}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "bleu": 0.8782107922177782, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.06233908637472536}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 0, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_1.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_precision": 0.12393482340516937, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0018630518464727952}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_recall": 0.12238578809900262, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0018120517865065748}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_fmeasure": 0.10705300854288186, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0013462774714981252}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_precision": 0.010161661868693592, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0007419379747381215}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_recall": 0.009731696294403758, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0006337255740513875}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_fmeasure": 0.008108504268322806, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0004889200349477446}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_precision": 0.0992345774655696, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0014859307313855661}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_recall": 0.09807549674162262, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.001429728781085387}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_fmeasure": 0.0850694760836755, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.001006857986151386}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_precision": 0.1193707362486968, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.001785081125105243}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_recall": 0.11807674303541536, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0017325391179047619}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_fmeasure": 0.10318864443982732, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0012842295844517349}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "bleu": 0.6949458806207459, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.07714277370064401}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 1, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_2.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_precision": 0.19955841396066626, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.003563362564684984}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_recall": 0.17976730548874567, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.002812039034680942}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_fmeasure": 0.15154565635865028, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.002014193604109947}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_precision": 0.04636727493908201, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.002063674255259155}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_recall": 0.03608850885731014, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.001271336103808331}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_fmeasure": 0.030157548716390026, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.000979656882582934}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_precision": 0.15985580048274545, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0030656099486307295}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_recall": 0.13936147048042474, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0021345192396162484}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_fmeasure": 0.11730616110922598, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.001484351215540235}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_precision": 0.1888526606228872, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.003410798800001199}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_recall": 0.1691336034450824, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.002624858772477078}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_fmeasure": 0.1427247039190245, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0018804491933968692}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "bleu": 2.071533371517729, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.05897326993999797}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 2, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_3.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_precision": 0.20783910831244695, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0041056369118362915}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_recall": 0.15315652086868922, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0029219941762290836}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_fmeasure": 0.13728912554108974, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.002210636879037005}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_precision": 0.05270913707602405, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.002285548107515653}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_recall": 0.035017856600902064, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0012792022939514419}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_fmeasure": 0.030998867710424604, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0010349607233972873}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_precision": 0.1691865105741514, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.003526604807735842}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_recall": 0.11995321142575928, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.002237431144354683}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_fmeasure": 0.10784177879573671, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.001680754333177093}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_precision": 0.19613380117261478, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.003915706153071064}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_recall": 0.14354467078945166, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.002730148158202835}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_fmeasure": 0.12888832560899113, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.002065212581324766}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "bleu": 1.7158265250796738, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.08255493477628999}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 3, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_4.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_precision": 0.07489403465656087, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0031490707471580844}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_recall": 0.048110228553027476, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.002076564942166141}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_fmeasure": 0.045629902283143044, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0017825266978103052}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_precision": 0.020284440274304042, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0015803898910115802}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_recall": 0.01150753413486689, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0008049360289966782}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_fmeasure": 0.010857787835966816, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0006985632492924999}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_precision": 0.06174056483906975, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.00269329372354837}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_recall": 0.037833085579516874, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0016105357637920758}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_fmeasure": 0.03609966283150641, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0013906297088899055}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_precision": 0.07068041799795843, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0030038060995320245}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_recall": 0.044646360061268694, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0019119530057766043}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_fmeasure": 0.04240858030577188, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0016374264677706336}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "bleu": 0.03397931679687075, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.010219720048480601}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 4, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_rephrase_en_5.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_precision": 0.01267880786453634, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0014107718837642158}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_recall": 0.006829802784380045, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0007969086886020665}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge1_fmeasure": 0.007009996490132747, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0007483893135097644}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_precision": 0.0037730408348988047, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0006557006638120141}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_recall": 0.0016812158510089643, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.00028258655267791193}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rouge2_fmeasure": 0.001884907029887383, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0003106992853917735}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_precision": 0.010660159857783067, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0012137585306723504}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_recall": 0.005539711110344907, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0006404529653601033}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeL_fmeasure": 0.005745242124119561, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0006166861167801969}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_precision": 0.011917111047382597, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0013457069158883772}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_recall": 0.006387289133689629, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0007525453532945123}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "rougeLsum_fmeasure": 0.006514389001825069, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0006942350509360657}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "rephrase_en", "bleu": 1.436251918375344e-17, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "753f0a46-aeff-4cd2-932c-8548897cebe5", "prompt_jinja": "{{source}}\n\nHow would you rephrase that briefly in English? ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 2.518456925923897e-16}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 5, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_0.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_precision": 0.10827554139384789, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0018480483599203554}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_recall": 0.1601220824903473, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.002511761468967375}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_fmeasure": 0.11797375573232652, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.001787932549239138}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_precision": 0.016732009165892972, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0006411115552393211}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_recall": 0.025955852461074132, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0010372881852128855}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_fmeasure": 0.018337631069464823, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0006645410421914484}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_precision": 0.09051719966984409, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.001452514887421999}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_recall": 0.13515257158124425, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.002014484368668383}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_fmeasure": 0.09877513805132598, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.001378154670190176}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_precision": 0.09979611184326491, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0017007142899985582}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_recall": 0.14851440218798725, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0023436070686969116}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_fmeasure": 0.10887644572618904, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.00164463481571714}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "bleu": 0.8302747934412187, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.03099763809793827}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 0, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_1.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_precision": 0.14038255710904818, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0022362237278879147}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_recall": 0.1378058499912762, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0021035303193778104}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_fmeasure": 0.11955053789402524, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.001553594400301458}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_precision": 0.01604633886618366, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0009832238101183293}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_recall": 0.015105077973850791, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0008575214995645402}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_fmeasure": 0.0123860791723954, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0006234618720099309}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_precision": 0.11171129737580932, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0018144880458233971}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_recall": 0.10887363072310992, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0016129000451380686}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_fmeasure": 0.09389730283947277, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0011344181369359976}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_precision": 0.13462495654091725, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.002149688036909027}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_recall": 0.13200488540776753, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0019904127573549275}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_fmeasure": 0.11445063468833441, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0014620548924494364}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "bleu": 1.0143656082276766, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.06230806130294888}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 1, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_2.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_precision": 0.2033469599245341, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.003533234544383998}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_recall": 0.16241044106899688, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0024440756470575815}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_fmeasure": 0.1468801060545036, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.001912430275231926}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_precision": 0.0436491790167234, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0019690486829533837}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_recall": 0.029794904677061543, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0011192076043048031}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_fmeasure": 0.02707468555655967, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0009333721234606784}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_precision": 0.16308315193371065, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.002940788540223112}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_recall": 0.12878877169543396, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0019119698643611006}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_fmeasure": 0.11576625722676982, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.001446447991054302}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_precision": 0.19124816902577402, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0033385395250822964}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_recall": 0.15297611443406722, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.002300286276263116}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_fmeasure": 0.13804831106314366, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0017814784865369622}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "bleu": 1.7954754371153347, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.09925045071096054}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 2, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_3.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_precision": 0.19810601765284833, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0040826496375516086}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_recall": 0.13120505099089064, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0026391728313708033}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_fmeasure": 0.12467719071250928, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.002118628385855177}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_precision": 0.047093482566777735, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.002160792983592125}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_recall": 0.027456987679276578, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0011665569741212422}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_fmeasure": 0.025707366639969195, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0009651585388355819}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_precision": 0.16153236801178705, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.003445446855244506}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_recall": 0.10446745914677141, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0020640521394088853}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_fmeasure": 0.099109079823698, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.001619881250413107}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_precision": 0.18718878537995878, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.003894314509709987}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_recall": 0.12362189047751815, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.002483767198983207}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_fmeasure": 0.11738591907799324, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.00198575551575682}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "bleu": 1.4181139783319565, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.07949256570752532}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 3, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_4.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_precision": 0.0680705622852744, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0030507778424628115}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_recall": 0.038179860498366525, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.001718388118950408}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_fmeasure": 0.039578571113952944, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0016625131022348848}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_precision": 0.018644591523780108, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0015559914596925517}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_recall": 0.0080529351245613, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.000621424810142374}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_fmeasure": 0.008833712329002994, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0006567333321688669}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_precision": 0.05637347753592951, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0025959121030370578}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_recall": 0.030466339478315622, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0013578331379260548}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_fmeasure": 0.03177319397499743, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0013286760797165292}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_precision": 0.06470007785014378, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.002932588330383202}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_recall": 0.03569472206069653, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.001599322013259645}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_fmeasure": 0.037138138976286025, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0015541979170414365}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "bleu": 0.010213266445722532, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.003236546636321977}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 4, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_summarize_above_en_5.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_precision": 0.009276147164957583, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.0011516412401362686}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_recall": 0.0056634584391191505, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0007248275837093809}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge1_fmeasure": 0.0055973738841194965, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.000652708260644998}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_precision": 0.001943157597014003, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.00048461355635539036}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_recall": 0.000944134401646025, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.00019244455859641443}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rouge2_fmeasure": 0.0009844255793615558, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0001914471696270827}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_precision": 0.007719327921024845, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.000993107843358693}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_recall": 0.004493962145305658, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.000570203812219779}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeL_fmeasure": 0.0044738408838412005, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.000518658227497335}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_precision": 0.008599455993650213, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0010784542769918472}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_recall": 0.00523907271902791, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.000676678856255735}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "rougeLsum_fmeasure": 0.005160009360078556, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.000603835567495784}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "summarize_above_en", "bleu": 8.217999349564618e-20, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "088288f3-7516-4cf7-9406-0e082053bf54", "prompt_jinja": "{{source}}\n\n===\n\nWrite a summary of the text above in English : ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 4.589980483058814e-18}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 5, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_tldr_en_0.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge1_precision": 0.15231044332305022, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.001979330420580656}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge1_recall": 0.2396531912121356, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.002552865784270149}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge1_fmeasure": 0.17249271064295765, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.0018543316864917051}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge2_precision": 0.029711681766300943, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0007971081048451994}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge2_recall": 0.04835172980239146, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0013203369767684852}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge2_fmeasure": 0.03380930288340197, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.0008466232007854189}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeL_precision": 0.12414165678595113, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0014463847617503154}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeL_recall": 0.20174092478712785, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0021196350776473733}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeL_fmeasure": 0.14223243736652721, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0013811850966333327}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeLsum_precision": 0.13814381149285762, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.0017872142639755628}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeLsum_recall": 0.2184991871149405, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.0023376496412298928}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeLsum_fmeasure": 0.15673609637062036, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.001680282283820491}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "bleu": 1.4528250679055472, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.07772013054387669}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 0, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}
2b246b46b/eval/agg.lm1-2b2-46b_GEM-wiki_lingua_en_tldr_en_1.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"results": [{"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge1_precision": 0.200162872123937, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_precision_stderr": 0.003336407093344552}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge1_recall": 0.18849805250080198, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_recall_stderr": 0.0027760507322805343}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge1_fmeasure": 0.1602577791225178, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge1_fmeasure_stderr": 0.002026542297238986}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge2_precision": 0.0428643976744321, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_precision_stderr": 0.0018768572537315072}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge2_recall": 0.035820675595350786, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_recall_stderr": 0.0012701573089832152}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rouge2_fmeasure": 0.03044802408852658, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rouge2_fmeasure_stderr": 0.001043083619557999}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeL_precision": 0.15601004633581056, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_precision_stderr": 0.0027909069669524774}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeL_recall": 0.14290940740640337, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_recall_stderr": 0.0020578021246362922}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeL_fmeasure": 0.121649238934672, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeL_fmeasure_stderr": 0.0014911189279727588}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeLsum_precision": 0.18912189875595237, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_precision_stderr": 0.003171839708129723}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeLsum_recall": 0.1776337478196078, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_recall_stderr": 0.002588079165595503}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "rougeLsum_fmeasure": 0.15101682874720862, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "rougeLsum_fmeasure_stderr": 0.0018796005080787329}, {"task_name": "GEM/wiki_lingua_en", "prompt_name": "tldr_en", "bleu": 2.0509008703876277, "fixed_answer_choice_list": null, "dataset_path": "GEM/wiki_lingua", "dataset_name": "en", "subset": null, "prompt_id": "d3c5baa3-5e37-46f8-b1b2-5b834181c9da", "prompt_jinja": "{{source}}\n\nTL;DR in English: ||| {{target}}", "prompt_original_task": true, "comment": "", "bleu_stderr": 0.09414082291927084}], "config": {"model": "hf-causal", "model_args": "pretrained=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/lm1-2b2-46b/transformers,use_accelerate=True,tokenizer=/pfs/lustrep4/scratch/project_462000119/muennighoff/nov-2022-bettercom/gpt2,dtype=bfloat16", "task_args": "", "num_fewshot": 1, "batch_size": 16, "device": "cuda", "use_cache": false, "limit": 3000, "bootstrap_iters": 10, "seed": 1234}}