bigscience/bloomz-7b1-p3
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: bigscience/bloomz-7b1-p3
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-06-15 07:40:14 +08:00
commit 78a6661ff1
634 changed files with 7477 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ar/GPT-3_style_armt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ar",
"template_name": "GPT-3 style_armt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='GPT-3 style_armt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ar/MNLI_crowdsource_armt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ar",
"template_name": "MNLI crowdsource_armt",
"evaluation": {
"accuracy": 0.42891566265060244
},
"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='MNLI crowdsource_armt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ar/can_we_infer_armt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ar",
"template_name": "can we infer_armt",
"evaluation": {
"accuracy": 0.3353413654618474
},
"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='can we infer_armt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ar/guaranteed_possible_impossible_armt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ar",
"template_name": "guaranteed/possible/impossible_armt",
"evaluation": {
"accuracy": 0.3755020080321285
},
"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='guaranteed/possible/impossible_armt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ar/justified_in_saying_armt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ar",
"template_name": "justified in saying_armt",
"evaluation": {
"accuracy": 0.3349397590361446
},
"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ar', template_name='justified in saying_armt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/es/GPT-3_style_esmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "es",
"template_name": "GPT-3 style_esmt",
"evaluation": {
"accuracy": 0.5220883534136547
},
"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='GPT-3 style_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/es/MNLI_crowdsource_esmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "es",
"template_name": "MNLI crowdsource_esmt",
"evaluation": {
"accuracy": 0.4847389558232932
},
"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='MNLI crowdsource_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/es/can_we_infer_esmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "es",
"template_name": "can we infer_esmt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='can we infer_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/es/guaranteed_possible_impossible_esmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "es",
"template_name": "guaranteed/possible/impossible_esmt",
"evaluation": {
"accuracy": 0.3449799196787149
},
"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='guaranteed/possible/impossible_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/es/justified_in_saying_esmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "es",
"template_name": "justified in saying_esmt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='es', template_name='justified in saying_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/fr/GPT-3_style_frmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "fr",
"template_name": "GPT-3 style_frmt",
"evaluation": {
"accuracy": 0.4791164658634538
},
"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='GPT-3 style_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/fr/MNLI_crowdsource_frmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "fr",
"template_name": "MNLI crowdsource_frmt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='MNLI crowdsource_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/fr/can_we_infer_frmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "fr",
"template_name": "can we infer_frmt",
"evaluation": {
"accuracy": 0.42248995983935744
},
"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='can we infer_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/fr/guaranteed_possible_impossible_frmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "fr",
"template_name": "guaranteed/possible/impossible_frmt",
"evaluation": {
"accuracy": 0.41847389558232934
},
"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='guaranteed/possible/impossible_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/fr/justified_in_saying_frmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "fr",
"template_name": "justified in saying_frmt",
"evaluation": {
"accuracy": 0.378714859437751
},
"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='fr', template_name='justified in saying_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/hi/GPT-3_style_himt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "hi",
"template_name": "GPT-3 style_himt",
"evaluation": {
"accuracy": 0.3389558232931727
},
"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='GPT-3 style_himt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/hi/MNLI_crowdsource_himt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "hi",
"template_name": "MNLI crowdsource_himt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='MNLI crowdsource_himt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/hi/can_we_infer_himt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "hi",
"template_name": "can we infer_himt",
"evaluation": {
"accuracy": 0.3542168674698795
},
"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='can we infer_himt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/hi/guaranteed_possible_impossible_himt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "hi",
"template_name": "guaranteed/possible/impossible_himt",
"evaluation": {
"accuracy": 0.3353413654618474
},
"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='guaranteed/possible/impossible_himt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/hi/justified_in_saying_himt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "hi",
"template_name": "justified in saying_himt",
"evaluation": {
"accuracy": 0.39879518072289155
},
"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='hi', template_name='justified in saying_himt', tokenizer_name=None, use_slow_tokenizer=False)"
}

50
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/merged.csv Normal file
View File

@@ -0,0 +1,50 @@
dataset,prompt,metric,value
xnli_ar,GPT-3 style_armt,accuracy,0.3333333333333333
xnli_ar,MNLI crowdsource_armt,accuracy,0.42891566265060244
xnli_ar,can we infer_armt,accuracy,0.3353413654618474
xnli_ar,guaranteed/possible/impossible_armt,accuracy,0.3755020080321285
xnli_ar,justified in saying_armt,accuracy,0.3349397590361446
xnli_ar,median,accuracy,0.3353413654618474
xnli_es,GPT-3 style_esmt,accuracy,0.5220883534136547
xnli_es,MNLI crowdsource_esmt,accuracy,0.4847389558232932
xnli_es,can we infer_esmt,accuracy,0.3333333333333333
xnli_es,guaranteed/possible/impossible_esmt,accuracy,0.3449799196787149
xnli_es,justified in saying_esmt,accuracy,0.3333333333333333
xnli_es,median,accuracy,0.3449799196787149
xnli_fr,GPT-3 style_frmt,accuracy,0.4791164658634538
xnli_fr,MNLI crowdsource_frmt,accuracy,0.3333333333333333
xnli_fr,can we infer_frmt,accuracy,0.42248995983935744
xnli_fr,guaranteed/possible/impossible_frmt,accuracy,0.41847389558232934
xnli_fr,justified in saying_frmt,accuracy,0.378714859437751
xnli_fr,median,accuracy,0.41847389558232934
xnli_hi,GPT-3 style_himt,accuracy,0.3389558232931727
xnli_hi,MNLI crowdsource_himt,accuracy,0.3333333333333333
xnli_hi,can we infer_himt,accuracy,0.3542168674698795
xnli_hi,guaranteed/possible/impossible_himt,accuracy,0.3353413654618474
xnli_hi,justified in saying_himt,accuracy,0.39879518072289155
xnli_hi,median,accuracy,0.3389558232931727
xnli_sw,GPT-3 style_swmt,accuracy,0.3333333333333333
xnli_sw,MNLI crowdsource_swmt,accuracy,0.3333333333333333
xnli_sw,can we infer_swmt,accuracy,0.334136546184739
xnli_sw,guaranteed/possible/impossible_swmt,accuracy,0.3236947791164659
xnli_sw,justified in saying_swmt,accuracy,0.3321285140562249
xnli_sw,median,accuracy,0.3333333333333333
xnli_ur,GPT-3 style_urmt,accuracy,0.3751004016064257
xnli_ur,MNLI crowdsource_urmt,accuracy,0.3751004016064257
xnli_ur,can we infer_urmt,accuracy,0.329718875502008
xnli_ur,guaranteed/possible/impossible_urmt,accuracy,0.3337349397590361
xnli_ur,justified in saying_urmt,accuracy,0.3285140562248996
xnli_ur,median,accuracy,0.3337349397590361
xnli_vi,GPT-3 style_vimt,accuracy,0.3333333333333333
xnli_vi,MNLI crowdsource_vimt,accuracy,0.3333333333333333
xnli_vi,can we infer_vimt,accuracy,0.342570281124498
xnli_vi,guaranteed/possible/impossible_vimt,accuracy,0.3333333333333333
xnli_vi,justified in saying_vimt,accuracy,0.3365461847389558
xnli_vi,median,accuracy,0.3333333333333333
xnli_zh,GPT-3 style_zhmt,accuracy,0.3606425702811245
xnli_zh,MNLI crowdsource_zhmt,accuracy,0.39598393574297186
xnli_zh,can we infer_zhmt,accuracy,0.351004016064257
xnli_zh,guaranteed/possible/impossible_zhmt,accuracy,0.3473895582329317
xnli_zh,justified in saying_zhmt,accuracy,0.3409638554216867
xnli_zh,median,accuracy,0.351004016064257
multiple,average,multiple,0.348644578313253
1 dataset prompt metric value
2 xnli_ar GPT-3 style_armt accuracy 0.3333333333333333
3 xnli_ar MNLI crowdsource_armt accuracy 0.42891566265060244
4 xnli_ar can we infer_armt accuracy 0.3353413654618474
5 xnli_ar guaranteed/possible/impossible_armt accuracy 0.3755020080321285
6 xnli_ar justified in saying_armt accuracy 0.3349397590361446
7 xnli_ar median accuracy 0.3353413654618474
8 xnli_es GPT-3 style_esmt accuracy 0.5220883534136547
9 xnli_es MNLI crowdsource_esmt accuracy 0.4847389558232932
10 xnli_es can we infer_esmt accuracy 0.3333333333333333
11 xnli_es guaranteed/possible/impossible_esmt accuracy 0.3449799196787149
12 xnli_es justified in saying_esmt accuracy 0.3333333333333333
13 xnli_es median accuracy 0.3449799196787149
14 xnli_fr GPT-3 style_frmt accuracy 0.4791164658634538
15 xnli_fr MNLI crowdsource_frmt accuracy 0.3333333333333333
16 xnli_fr can we infer_frmt accuracy 0.42248995983935744
17 xnli_fr guaranteed/possible/impossible_frmt accuracy 0.41847389558232934
18 xnli_fr justified in saying_frmt accuracy 0.378714859437751
19 xnli_fr median accuracy 0.41847389558232934
20 xnli_hi GPT-3 style_himt accuracy 0.3389558232931727
21 xnli_hi MNLI crowdsource_himt accuracy 0.3333333333333333
22 xnli_hi can we infer_himt accuracy 0.3542168674698795
23 xnli_hi guaranteed/possible/impossible_himt accuracy 0.3353413654618474
24 xnli_hi justified in saying_himt accuracy 0.39879518072289155
25 xnli_hi median accuracy 0.3389558232931727
26 xnli_sw GPT-3 style_swmt accuracy 0.3333333333333333
27 xnli_sw MNLI crowdsource_swmt accuracy 0.3333333333333333
28 xnli_sw can we infer_swmt accuracy 0.334136546184739
29 xnli_sw guaranteed/possible/impossible_swmt accuracy 0.3236947791164659
30 xnli_sw justified in saying_swmt accuracy 0.3321285140562249
31 xnli_sw median accuracy 0.3333333333333333
32 xnli_ur GPT-3 style_urmt accuracy 0.3751004016064257
33 xnli_ur MNLI crowdsource_urmt accuracy 0.3751004016064257
34 xnli_ur can we infer_urmt accuracy 0.329718875502008
35 xnli_ur guaranteed/possible/impossible_urmt accuracy 0.3337349397590361
36 xnli_ur justified in saying_urmt accuracy 0.3285140562248996
37 xnli_ur median accuracy 0.3337349397590361
38 xnli_vi GPT-3 style_vimt accuracy 0.3333333333333333
39 xnli_vi MNLI crowdsource_vimt accuracy 0.3333333333333333
40 xnli_vi can we infer_vimt accuracy 0.342570281124498
41 xnli_vi guaranteed/possible/impossible_vimt accuracy 0.3333333333333333
42 xnli_vi justified in saying_vimt accuracy 0.3365461847389558
43 xnli_vi median accuracy 0.3333333333333333
44 xnli_zh GPT-3 style_zhmt accuracy 0.3606425702811245
45 xnli_zh MNLI crowdsource_zhmt accuracy 0.39598393574297186
46 xnli_zh can we infer_zhmt accuracy 0.351004016064257
47 xnli_zh guaranteed/possible/impossible_zhmt accuracy 0.3473895582329317
48 xnli_zh justified in saying_zhmt accuracy 0.3409638554216867
49 xnli_zh median accuracy 0.351004016064257
50 multiple average multiple 0.348644578313253

1
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/merged.json Normal file
View File

File diff suppressed because one or more lines are too long

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/sw/GPT-3_style_swmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "sw",
"template_name": "GPT-3 style_swmt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='GPT-3 style_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/sw/MNLI_crowdsource_swmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "sw",
"template_name": "MNLI crowdsource_swmt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='MNLI crowdsource_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/sw/can_we_infer_swmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "sw",
"template_name": "can we infer_swmt",
"evaluation": {
"accuracy": 0.334136546184739
},
"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='can we infer_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/sw/guaranteed_possible_impossible_swmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "sw",
"template_name": "guaranteed/possible/impossible_swmt",
"evaluation": {
"accuracy": 0.3236947791164659
},
"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='guaranteed/possible/impossible_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/sw/justified_in_saying_swmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "sw",
"template_name": "justified in saying_swmt",
"evaluation": {
"accuracy": 0.3321285140562249
},
"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='sw', template_name='justified in saying_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ur/GPT-3_style_urmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ur",
"template_name": "GPT-3 style_urmt",
"evaluation": {
"accuracy": 0.3751004016064257
},
"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='GPT-3 style_urmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ur/MNLI_crowdsource_urmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ur",
"template_name": "MNLI crowdsource_urmt",
"evaluation": {
"accuracy": 0.3751004016064257
},
"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='MNLI crowdsource_urmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ur/can_we_infer_urmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ur",
"template_name": "can we infer_urmt",
"evaluation": {
"accuracy": 0.329718875502008
},
"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='can we infer_urmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ur/guaranteed_possible_impossible_urmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ur",
"template_name": "guaranteed/possible/impossible_urmt",
"evaluation": {
"accuracy": 0.3337349397590361
},
"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='guaranteed/possible/impossible_urmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/ur/justified_in_saying_urmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "ur",
"template_name": "justified in saying_urmt",
"evaluation": {
"accuracy": 0.3285140562248996
},
"arguments": "Namespace(config_name=None, dataset_config_name='ur', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ur', template_name='justified in saying_urmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/vi/GPT-3_style_vimt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "vi",
"template_name": "GPT-3 style_vimt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='GPT-3 style_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/vi/MNLI_crowdsource_vimt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "vi",
"template_name": "MNLI crowdsource_vimt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='MNLI crowdsource_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/vi/can_we_infer_vimt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "vi",
"template_name": "can we infer_vimt",
"evaluation": {
"accuracy": 0.342570281124498
},
"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='can we infer_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/vi/guaranteed_possible_impossible_vimt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "vi",
"template_name": "guaranteed/possible/impossible_vimt",
"evaluation": {
"accuracy": 0.3333333333333333
},
"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='guaranteed/possible/impossible_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/vi/justified_in_saying_vimt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "vi",
"template_name": "justified in saying_vimt",
"evaluation": {
"accuracy": 0.3365461847389558
},
"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='vi', template_name='justified in saying_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/zh/GPT-3_style_zhmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "zh",
"template_name": "GPT-3 style_zhmt",
"evaluation": {
"accuracy": 0.3606425702811245
},
"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='GPT-3 style_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/zh/MNLI_crowdsource_zhmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "zh",
"template_name": "MNLI crowdsource_zhmt",
"evaluation": {
"accuracy": 0.39598393574297186
},
"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='MNLI crowdsource_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/zh/can_we_infer_zhmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "zh",
"template_name": "can we infer_zhmt",
"evaluation": {
"accuracy": 0.351004016064257
},
"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='can we infer_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/zh/guaranteed_possible_impossible_zhmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "zh",
"template_name": "guaranteed/possible/impossible_zhmt",
"evaluation": {
"accuracy": 0.3473895582329317
},
"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='guaranteed/possible/impossible_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
}

9
evaluation_bloomz-7b1-p3/evaluation_xnlimt/xnli/zh/justified_in_saying_zhmt/results.json Normal file
View File

@@ -0,0 +1,9 @@
{
"dataset_name": "xnli",
"dataset_config_name": "zh",
"template_name": "justified in saying_zhmt",
"evaluation": {
"accuracy": 0.3409638554216867
},
"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/p31lossseqglobal_step1000/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='zh', template_name='justified in saying_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
}