diff --git a/.gitattributes b/.gitattributes
index 637fa167e56685c01bc97f08a420ea76330cf6df..1fd4826e6f1aafb9303f7a6f9709083bd5723fc3 100644
--- a/.gitattributes
+++ b/.gitattributes
@@ -30,3 +30,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text
diff --git a/config.json b/config.json
new file mode 100644
index 0000000000000000000000000000000000000000..faafcee93060917fc6c1428a7e986b02e7496c66
--- /dev/null
+++ b/config.json
@@ -0,0 +1,31 @@
+{
+  "apply_residual_connection_post_layernorm": false,
+  "architectures": [
+    "BloomModel"
+  ],
+  "attention_dropout": 0.0,
+  "attention_softmax_in_fp32": true,
+  "bias_dropout_fusion": true,
+  "bos_token_id": 1,
+  "eos_token_id": 2,
+  "hidden_dropout": 0.0,
+  "initializer_range": 0.02,
+  "layer_norm_epsilon": 1e-05,
+  "masked_softmax_fusion": true,
+  "model_type": "bloom",
+  "n_embed": 4096,
+  "n_inner": null,
+  "n_layer": 30,
+  "num_attention_heads": 32,
+  "offset_alibi": 100,
+  "pad_token_id": 3,
+  "pretraining_tp": 4,
+  "seq_length": 2048,
+  "skip_bias_add": true,
+  "skip_bias_add_qkv": false,
+  "slow_but_exact": false,
+  "transformers_version": "4.21.0.dev0",
+  "unk_token_id": 0,
+  "use_cache": true,
+  "vocab_size": 250880
+}
diff --git a/evaluation/xnliht/ar/GPT-3_style_arht/results.json b/evaluation/xnliht/ar/GPT-3_style_arht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..89e43bafb8376a00e9550fe6e6650f8fa643b203
--- /dev/null
+++ b/evaluation/xnliht/ar/GPT-3_style_arht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "GPT-3 style_arht",
+  "evaluation": {
+    "accuracy": 0.4610441767068273
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/ar/MNLI_crowdsource_arht/results.json b/evaluation/xnliht/ar/MNLI_crowdsource_arht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..9f61804d9d11f714cbd249fc74308c8fbc0699c3
--- /dev/null
+++ b/evaluation/xnliht/ar/MNLI_crowdsource_arht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "MNLI crowdsource_arht",
+  "evaluation": {
+    "accuracy": 0.3899598393574297
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/ar/can_we_infer_arht/results.json b/evaluation/xnliht/ar/can_we_infer_arht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..390cd9f97ace2cb2f965cf64641250954c6e21fb
--- /dev/null
+++ b/evaluation/xnliht/ar/can_we_infer_arht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "can we infer_arht",
+  "evaluation": {
+    "accuracy": 0.3550200803212851
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/ar/guaranteed_possible_impossible_arht/results.json b/evaluation/xnliht/ar/guaranteed_possible_impossible_arht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b6e2d83aadea0aca86a997c8bf99a61c2ab7063d
--- /dev/null
+++ b/evaluation/xnliht/ar/guaranteed_possible_impossible_arht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "guaranteed/possible/impossible_arht",
+  "evaluation": {
+    "accuracy": 0.45461847389558235
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/ar/justified_in_saying_arht/results.json b/evaluation/xnliht/ar/justified_in_saying_arht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..82e207df2d935a573bad6b7501d6aaf42dcc1cb3
--- /dev/null
+++ b/evaluation/xnliht/ar/justified_in_saying_arht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "justified in saying_arht",
+  "evaluation": {
+    "accuracy": 0.3538152610441767
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/es/GPT-3_style_esht/results.json b/evaluation/xnliht/es/GPT-3_style_esht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..cffee59d56f1ebbb02db7d7721ae08ba1a4cddc8
--- /dev/null
+++ b/evaluation/xnliht/es/GPT-3_style_esht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "GPT-3 style_esht",
+  "evaluation": {
+    "accuracy": 0.5313253012048192
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/es/MNLI_crowdsource_esht/results.json b/evaluation/xnliht/es/MNLI_crowdsource_esht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..49638017ffc7ea741445cda68c9f802078b835d9
--- /dev/null
+++ b/evaluation/xnliht/es/MNLI_crowdsource_esht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "MNLI crowdsource_esht",
+  "evaluation": {
+    "accuracy": 0.334136546184739
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/es/can_we_infer_esht/results.json b/evaluation/xnliht/es/can_we_infer_esht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..de94c4c56f39e025ea75d755f6e27e9137a4bee7
--- /dev/null
+++ b/evaluation/xnliht/es/can_we_infer_esht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "can we infer_esht",
+  "evaluation": {
+    "accuracy": 0.36987951807228914
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/es/guaranteed_possible_impossible_esht/results.json b/evaluation/xnliht/es/guaranteed_possible_impossible_esht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..4ab106672cdff63d80d3a3f17628c4df8299825c
--- /dev/null
+++ b/evaluation/xnliht/es/guaranteed_possible_impossible_esht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "guaranteed/possible/impossible_esht",
+  "evaluation": {
+    "accuracy": 0.4686746987951807
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/es/justified_in_saying_esht/results.json b/evaluation/xnliht/es/justified_in_saying_esht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e1a78271a48bb847416de3e695e9ce8b9b0313e
--- /dev/null
+++ b/evaluation/xnliht/es/justified_in_saying_esht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "justified in saying_esht",
+  "evaluation": {
+    "accuracy": 0.37630522088353413
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/fr/GPT-3_style_frht/results.json b/evaluation/xnliht/fr/GPT-3_style_frht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee71295e8d994debecfdb9ef2012121716cfbcae
--- /dev/null
+++ b/evaluation/xnliht/fr/GPT-3_style_frht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "GPT-3 style_frht",
+  "evaluation": {
+    "accuracy": 0.5345381526104418
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/fr/MNLI_crowdsource_frht/results.json b/evaluation/xnliht/fr/MNLI_crowdsource_frht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7a42eef9058f181bac68d4804d7eb2e9f70fe4d6
--- /dev/null
+++ b/evaluation/xnliht/fr/MNLI_crowdsource_frht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "MNLI crowdsource_frht",
+  "evaluation": {
+    "accuracy": 0.3357429718875502
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/fr/can_we_infer_frht/results.json b/evaluation/xnliht/fr/can_we_infer_frht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..90a19abe8e6077dacc39c17c9bc153c75706f052
--- /dev/null
+++ b/evaluation/xnliht/fr/can_we_infer_frht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "can we infer_frht",
+  "evaluation": {
+    "accuracy": 0.5224899598393574
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/fr/guaranteed_possible_impossible_frht/results.json b/evaluation/xnliht/fr/guaranteed_possible_impossible_frht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1c3df1fa5dac930fd4024d0cf9f26589a2ec3188
--- /dev/null
+++ b/evaluation/xnliht/fr/guaranteed_possible_impossible_frht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "guaranteed/possible/impossible_frht",
+  "evaluation": {
+    "accuracy": 0.46586345381526106
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/fr/justified_in_saying_frht/results.json b/evaluation/xnliht/fr/justified_in_saying_frht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2bfb1f45fb8d49e5c2b459fdd5d3d8faefd65832
--- /dev/null
+++ b/evaluation/xnliht/fr/justified_in_saying_frht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "justified in saying_frht",
+  "evaluation": {
+    "accuracy": 0.4891566265060241
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/hi/GPT-3_style_hiht/results.json b/evaluation/xnliht/hi/GPT-3_style_hiht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..81126b8423ccfb2a338316e6d739e62f5b2ad908
--- /dev/null
+++ b/evaluation/xnliht/hi/GPT-3_style_hiht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "GPT-3 style_hiht",
+  "evaluation": {
+    "accuracy": 0.3325301204819277
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/hi/MNLI_crowdsource_hiht/results.json b/evaluation/xnliht/hi/MNLI_crowdsource_hiht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..72cc6afcf061bd27160824802dc13eea9fc91e3e
--- /dev/null
+++ b/evaluation/xnliht/hi/MNLI_crowdsource_hiht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "MNLI crowdsource_hiht",
+  "evaluation": {
+    "accuracy": 0.470281124497992
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/hi/can_we_infer_hiht/results.json b/evaluation/xnliht/hi/can_we_infer_hiht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d7bac61e418793f962192a821514feaa7e84b689
--- /dev/null
+++ b/evaluation/xnliht/hi/can_we_infer_hiht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "can we infer_hiht",
+  "evaluation": {
+    "accuracy": 0.37309236947791163
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/hi/guaranteed_possible_impossible_hiht/results.json b/evaluation/xnliht/hi/guaranteed_possible_impossible_hiht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..edc1bf25e960593d049d046709a7306a683453d1
--- /dev/null
+++ b/evaluation/xnliht/hi/guaranteed_possible_impossible_hiht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "guaranteed/possible/impossible_hiht",
+  "evaluation": {
+    "accuracy": 0.3514056224899598
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/hi/justified_in_saying_hiht/results.json b/evaluation/xnliht/hi/justified_in_saying_hiht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c7b4bf0fd416f731a886fbef444cfd580229a52
--- /dev/null
+++ b/evaluation/xnliht/hi/justified_in_saying_hiht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "justified in saying_hiht",
+  "evaluation": {
+    "accuracy": 0.3746987951807229
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/merged.csv b/evaluation/xnliht/merged.csv
new file mode 100644
index 0000000000000000000000000000000000000000..63d9419c23b657c9221fbcd5ed16fbbc8b1ca096
--- /dev/null
+++ b/evaluation/xnliht/merged.csv
@@ -0,0 +1,50 @@
+dataset,prompt,metric,value
+xnli_ar,GPT-3 style_arht,accuracy,0.4610441767068273
+xnli_ar,MNLI crowdsource_arht,accuracy,0.3899598393574297
+xnli_ar,can we infer_arht,accuracy,0.3550200803212851
+xnli_ar,guaranteed/possible/impossible_arht,accuracy,0.45461847389558235
+xnli_ar,justified in saying_arht,accuracy,0.3538152610441767
+xnli_ar,median,accuracy,0.3899598393574297
+xnli_es,GPT-3 style_esht,accuracy,0.5313253012048192
+xnli_es,MNLI crowdsource_esht,accuracy,0.334136546184739
+xnli_es,can we infer_esht,accuracy,0.36987951807228914
+xnli_es,guaranteed/possible/impossible_esht,accuracy,0.4686746987951807
+xnli_es,justified in saying_esht,accuracy,0.37630522088353413
+xnli_es,median,accuracy,0.37630522088353413
+xnli_fr,GPT-3 style_frht,accuracy,0.5345381526104418
+xnli_fr,MNLI crowdsource_frht,accuracy,0.3357429718875502
+xnli_fr,can we infer_frht,accuracy,0.5224899598393574
+xnli_fr,guaranteed/possible/impossible_frht,accuracy,0.46586345381526106
+xnli_fr,justified in saying_frht,accuracy,0.4891566265060241
+xnli_fr,median,accuracy,0.4891566265060241
+xnli_hi,GPT-3 style_hiht,accuracy,0.3325301204819277
+xnli_hi,MNLI crowdsource_hiht,accuracy,0.470281124497992
+xnli_hi,can we infer_hiht,accuracy,0.37309236947791163
+xnli_hi,guaranteed/possible/impossible_hiht,accuracy,0.3514056224899598
+xnli_hi,justified in saying_hiht,accuracy,0.3746987951807229
+xnli_hi,median,accuracy,0.37309236947791163
+xnli_sw,GPT-3 style_swht,accuracy,0.336144578313253
+xnli_sw,MNLI crowdsource_swht,accuracy,0.3333333333333333
+xnli_sw,can we infer_swht,accuracy,0.3453815261044177
+xnli_sw,guaranteed/possible/impossible_swht,accuracy,0.35582329317269074
+xnli_sw,justified in saying_swht,accuracy,0.3269076305220884
+xnli_sw,median,accuracy,0.336144578313253
+xnli_ur,GPT-3 style_urht,accuracy,0.4
+xnli_ur,MNLI crowdsource_urht,accuracy,0.3562248995983936
+xnli_ur,can we infer_urht,accuracy,0.3349397590361446
+xnli_ur,guaranteed/possible/impossible_urht,accuracy,0.37630522088353413
+xnli_ur,justified in saying_urht,accuracy,0.3405622489959839
+xnli_ur,median,accuracy,0.3562248995983936
+xnli_vi,GPT-3 style_viht,accuracy,0.5265060240963856
+xnli_vi,MNLI crowdsource_viht,accuracy,0.37710843373493974
+xnli_vi,can we infer_viht,accuracy,0.5116465863453815
+xnli_vi,guaranteed/possible/impossible_viht,accuracy,0.3578313253012048
+xnli_vi,justified in saying_viht,accuracy,0.5028112449799197
+xnli_vi,median,accuracy,0.5028112449799197
+xnli_zh,GPT-3 style_zhht,accuracy,0.3196787148594378
+xnli_zh,MNLI crowdsource_zhht,accuracy,0.38112449799196785
+xnli_zh,can we infer_zhht,accuracy,0.40642570281124496
+xnli_zh,guaranteed/possible/impossible_zhht,accuracy,0.344578313253012
+xnli_zh,justified in saying_zhht,accuracy,0.3369477911646586
+xnli_zh,median,accuracy,0.344578313253012
+multiple,average,multiple,0.39603413654618475
diff --git a/evaluation/xnliht/merged.json b/evaluation/xnliht/merged.json
new file mode 100644
index 0000000000000000000000000000000000000000..efc985e13b0d05e58fa83cc086ba7da085ef4275
--- /dev/null
+++ b/evaluation/xnliht/merged.json
@@ -0,0 +1 @@
+{"xnli_ar": {"GPT-3 style_arht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4610441767068273}, "template_name": "GPT-3 style_arht"}, "MNLI crowdsource_arht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3899598393574297}, "template_name": "MNLI crowdsource_arht"}, "can we infer_arht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3550200803212851}, "template_name": "can we infer_arht"}, "guaranteed/possible/impossible_arht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.45461847389558235}, "template_name": "guaranteed/possible/impossible_arht"}, "justified in saying_arht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_arht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3538152610441767}, "template_name": "justified in saying_arht"}}, "xnli_es": {"GPT-3 style_esht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5313253012048192}, "template_name": "GPT-3 style_esht"}, "MNLI crowdsource_esht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.334136546184739}, "template_name": "MNLI crowdsource_esht"}, "can we infer_esht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.36987951807228914}, "template_name": "can we infer_esht"}, "guaranteed/possible/impossible_esht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4686746987951807}, "template_name": "guaranteed/possible/impossible_esht"}, "justified in saying_esht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_esht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.37630522088353413}, "template_name": "justified in saying_esht"}}, "xnli_fr": {"GPT-3 style_frht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5345381526104418}, "template_name": "GPT-3 style_frht"}, "MNLI crowdsource_frht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3357429718875502}, "template_name": "MNLI crowdsource_frht"}, "can we infer_frht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5224899598393574}, "template_name": "can we infer_frht"}, "guaranteed/possible/impossible_frht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.46586345381526106}, "template_name": "guaranteed/possible/impossible_frht"}, "justified in saying_frht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_frht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4891566265060241}, "template_name": "justified in saying_frht"}}, "xnli_hi": {"GPT-3 style_hiht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3325301204819277}, "template_name": "GPT-3 style_hiht"}, "MNLI crowdsource_hiht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.470281124497992}, "template_name": "MNLI crowdsource_hiht"}, "can we infer_hiht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.37309236947791163}, "template_name": "can we infer_hiht"}, "guaranteed/possible/impossible_hiht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3514056224899598}, "template_name": "guaranteed/possible/impossible_hiht"}, "justified in saying_hiht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_hiht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3746987951807229}, "template_name": "justified in saying_hiht"}}, "xnli_sw": {"GPT-3 style_swht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.336144578313253}, "template_name": "GPT-3 style_swht"}, "MNLI crowdsource_swht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "MNLI crowdsource_swht"}, "can we infer_swht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3453815261044177}, "template_name": "can we infer_swht"}, "guaranteed/possible/impossible_swht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.35582329317269074}, "template_name": "guaranteed/possible/impossible_swht"}, "justified in saying_swht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3269076305220884}, "template_name": "justified in saying_swht"}}, "xnli_ur": {"GPT-3 style_urht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4}, "template_name": "GPT-3 style_urht"}, "MNLI crowdsource_urht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3562248995983936}, "template_name": "MNLI crowdsource_urht"}, "can we infer_urht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3349397590361446}, "template_name": "can we infer_urht"}, "guaranteed/possible/impossible_urht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.37630522088353413}, "template_name": "guaranteed/possible/impossible_urht"}, "justified in saying_urht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3405622489959839}, "template_name": "justified in saying_urht"}}, "xnli_vi": {"GPT-3 style_viht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5265060240963856}, "template_name": "GPT-3 style_viht"}, "MNLI crowdsource_viht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.37710843373493974}, "template_name": "MNLI crowdsource_viht"}, "can we infer_viht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5116465863453815}, "template_name": "can we infer_viht"}, "guaranteed/possible/impossible_viht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3578313253012048}, "template_name": "guaranteed/possible/impossible_viht"}, "justified in saying_viht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5028112449799197}, "template_name": "justified in saying_viht"}}, "xnli_zh": {"GPT-3 style_zhht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3196787148594378}, "template_name": "GPT-3 style_zhht"}, "MNLI crowdsource_zhht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.38112449799196785}, "template_name": "MNLI crowdsource_zhht"}, "can we infer_zhht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40642570281124496}, "template_name": "can we infer_zhht"}, "guaranteed/possible/impossible_zhht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.344578313253012}, "template_name": "guaranteed/possible/impossible_zhht"}, "justified in saying_zhht": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3369477911646586}, "template_name": "justified in saying_zhht"}}}
\ No newline at end of file
diff --git a/evaluation/xnliht/sw/GPT-3_style_swht/results.json b/evaluation/xnliht/sw/GPT-3_style_swht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..de666e20d45de89d7d2e0b87d2df2e5e636806a2
--- /dev/null
+++ b/evaluation/xnliht/sw/GPT-3_style_swht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "GPT-3 style_swht",
+  "evaluation": {
+    "accuracy": 0.336144578313253
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/sw/MNLI_crowdsource_swht/results.json b/evaluation/xnliht/sw/MNLI_crowdsource_swht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b299c5ccdc242196add1443520436e5e01b82db
--- /dev/null
+++ b/evaluation/xnliht/sw/MNLI_crowdsource_swht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "MNLI crowdsource_swht",
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/sw/can_we_infer_swht/results.json b/evaluation/xnliht/sw/can_we_infer_swht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b02921e07618dddbe9b0c9baa1bcf83488ffd879
--- /dev/null
+++ b/evaluation/xnliht/sw/can_we_infer_swht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "can we infer_swht",
+  "evaluation": {
+    "accuracy": 0.3453815261044177
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/sw/guaranteed_possible_impossible_swht/results.json b/evaluation/xnliht/sw/guaranteed_possible_impossible_swht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d0c7de87ad44d55d779ab418e11dc04eff4163a
--- /dev/null
+++ b/evaluation/xnliht/sw/guaranteed_possible_impossible_swht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "guaranteed/possible/impossible_swht",
+  "evaluation": {
+    "accuracy": 0.35582329317269074
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/sw/justified_in_saying_swht/results.json b/evaluation/xnliht/sw/justified_in_saying_swht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..09f98868f8a3f64df83efbfe2d61dc4507e3622e
--- /dev/null
+++ b/evaluation/xnliht/sw/justified_in_saying_swht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "justified in saying_swht",
+  "evaluation": {
+    "accuracy": 0.3269076305220884
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_swht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/ur/GPT-3_style_urht/results.json b/evaluation/xnliht/ur/GPT-3_style_urht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..cde9a7dc932d7fd11b347b05243ec4274928b418
--- /dev/null
+++ b/evaluation/xnliht/ur/GPT-3_style_urht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "GPT-3 style_urht",
+  "evaluation": {
+    "accuracy": 0.4
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/ur/MNLI_crowdsource_urht/results.json b/evaluation/xnliht/ur/MNLI_crowdsource_urht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a9c37fed7c6eaa7f42cae33a5ba7e0d82355619e
--- /dev/null
+++ b/evaluation/xnliht/ur/MNLI_crowdsource_urht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "MNLI crowdsource_urht",
+  "evaluation": {
+    "accuracy": 0.3562248995983936
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/ur/can_we_infer_urht/results.json b/evaluation/xnliht/ur/can_we_infer_urht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0275da97b071d9797bb41c979be0490b20c1a378
--- /dev/null
+++ b/evaluation/xnliht/ur/can_we_infer_urht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "can we infer_urht",
+  "evaluation": {
+    "accuracy": 0.3349397590361446
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/ur/guaranteed_possible_impossible_urht/results.json b/evaluation/xnliht/ur/guaranteed_possible_impossible_urht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..854f22010ac359d6fb9fd0586a760bdcfd66de48
--- /dev/null
+++ b/evaluation/xnliht/ur/guaranteed_possible_impossible_urht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "guaranteed/possible/impossible_urht",
+  "evaluation": {
+    "accuracy": 0.37630522088353413
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/ur/justified_in_saying_urht/results.json b/evaluation/xnliht/ur/justified_in_saying_urht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1b761bd5ca49c65720cd1d210425f00c303cb2ad
--- /dev/null
+++ b/evaluation/xnliht/ur/justified_in_saying_urht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "justified in saying_urht",
+  "evaluation": {
+    "accuracy": 0.3405622489959839
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_urht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/vi/GPT-3_style_viht/results.json b/evaluation/xnliht/vi/GPT-3_style_viht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..911fb94874ba7bd2ea8692a7c8021ac66b5153c2
--- /dev/null
+++ b/evaluation/xnliht/vi/GPT-3_style_viht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "GPT-3 style_viht",
+  "evaluation": {
+    "accuracy": 0.5265060240963856
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/vi/MNLI_crowdsource_viht/results.json b/evaluation/xnliht/vi/MNLI_crowdsource_viht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a896f3f9b9c16c00619928134371c5d8a06c8cf
--- /dev/null
+++ b/evaluation/xnliht/vi/MNLI_crowdsource_viht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "MNLI crowdsource_viht",
+  "evaluation": {
+    "accuracy": 0.37710843373493974
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/vi/can_we_infer_viht/results.json b/evaluation/xnliht/vi/can_we_infer_viht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..5592cf57ddff8a5624c29225b518325819692f3a
--- /dev/null
+++ b/evaluation/xnliht/vi/can_we_infer_viht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "can we infer_viht",
+  "evaluation": {
+    "accuracy": 0.5116465863453815
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/vi/guaranteed_possible_impossible_viht/results.json b/evaluation/xnliht/vi/guaranteed_possible_impossible_viht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..243d01383b0e661431abaf57e9da32a578e9728b
--- /dev/null
+++ b/evaluation/xnliht/vi/guaranteed_possible_impossible_viht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "guaranteed/possible/impossible_viht",
+  "evaluation": {
+    "accuracy": 0.3578313253012048
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/vi/justified_in_saying_viht/results.json b/evaluation/xnliht/vi/justified_in_saying_viht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d4eadded27b8fe4d524b10ca3bb45f007dfeca90
--- /dev/null
+++ b/evaluation/xnliht/vi/justified_in_saying_viht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "justified in saying_viht",
+  "evaluation": {
+    "accuracy": 0.5028112449799197
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_viht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/zh/GPT-3_style_zhht/results.json b/evaluation/xnliht/zh/GPT-3_style_zhht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d4254e656fdcba212a82643bc0e1902dc64567ff
--- /dev/null
+++ b/evaluation/xnliht/zh/GPT-3_style_zhht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "GPT-3 style_zhht",
+  "evaluation": {
+    "accuracy": 0.3196787148594378
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/zh/MNLI_crowdsource_zhht/results.json b/evaluation/xnliht/zh/MNLI_crowdsource_zhht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..10dbab5a4fe31be7999b7a95f3eee8e15829d6fb
--- /dev/null
+++ b/evaluation/xnliht/zh/MNLI_crowdsource_zhht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "MNLI crowdsource_zhht",
+  "evaluation": {
+    "accuracy": 0.38112449799196785
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/zh/can_we_infer_zhht/results.json b/evaluation/xnliht/zh/can_we_infer_zhht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..620f357300bfe4c9817e823269ac9f4e1f4f44a5
--- /dev/null
+++ b/evaluation/xnliht/zh/can_we_infer_zhht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "can we infer_zhht",
+  "evaluation": {
+    "accuracy": 0.40642570281124496
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/zh/guaranteed_possible_impossible_zhht/results.json b/evaluation/xnliht/zh/guaranteed_possible_impossible_zhht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..65f9e6a084a0d11b193fa456270487f87f7d9b80
--- /dev/null
+++ b/evaluation/xnliht/zh/guaranteed_possible_impossible_zhht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "guaranteed/possible/impossible_zhht",
+  "evaluation": {
+    "accuracy": 0.344578313253012
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnliht/zh/justified_in_saying_zhht/results.json b/evaluation/xnliht/zh/justified_in_saying_zhht/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1e999809f3f18ddd914a16705b305ea1a99d9038
--- /dev/null
+++ b/evaluation/xnliht/zh/justified_in_saying_zhht/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "justified in saying_zhht",
+  "evaluation": {
+    "accuracy": 0.3369477911646586
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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_zhht', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ar/GPT-3_style_armt/results.json b/evaluation/xnlimt/ar/GPT-3_style_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..77cfdf21c92824d790481216193934744f3571dd
--- /dev/null
+++ b/evaluation/xnlimt/ar/GPT-3_style_armt/results.json
@@ -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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ar/MNLI_crowdsource_armt/results.json b/evaluation/xnlimt/ar/MNLI_crowdsource_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..50233f2f7b48a56410b663a1e53062b48cb34970
--- /dev/null
+++ b/evaluation/xnlimt/ar/MNLI_crowdsource_armt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "MNLI crowdsource_armt",
+  "evaluation": {
+    "accuracy": 0.4855421686746988
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ar/can_we_infer_armt/results.json b/evaluation/xnlimt/ar/can_we_infer_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3c38307364767eeff17f0b7017a55ec6d59caccf
--- /dev/null
+++ b/evaluation/xnlimt/ar/can_we_infer_armt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "can we infer_armt",
+  "evaluation": {
+    "accuracy": 0.3413654618473896
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ar/guaranteed_possible_impossible_armt/results.json b/evaluation/xnlimt/ar/guaranteed_possible_impossible_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..62e5b9267ca8b55ec7427d8d404b4dc69e57c9ff
--- /dev/null
+++ b/evaluation/xnlimt/ar/guaranteed_possible_impossible_armt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "guaranteed/possible/impossible_armt",
+  "evaluation": {
+    "accuracy": 0.35542168674698793
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ar/justified_in_saying_armt/results.json b/evaluation/xnlimt/ar/justified_in_saying_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1f1f936f73b6370418470a32df93bc6ecc389d39
--- /dev/null
+++ b/evaluation/xnlimt/ar/justified_in_saying_armt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "justified in saying_armt",
+  "evaluation": {
+    "accuracy": 0.3465863453815261
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/es/GPT-3_style_esmt/results.json b/evaluation/xnlimt/es/GPT-3_style_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e1647e62c5110a1aa782d3f6bd3584976d02e177
--- /dev/null
+++ b/evaluation/xnlimt/es/GPT-3_style_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "GPT-3 style_esmt",
+  "evaluation": {
+    "accuracy": 0.5385542168674698
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/es/MNLI_crowdsource_esmt/results.json b/evaluation/xnlimt/es/MNLI_crowdsource_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..90b3f6eff9418f33f8581c2425f1323cbac2198a
--- /dev/null
+++ b/evaluation/xnlimt/es/MNLI_crowdsource_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "MNLI crowdsource_esmt",
+  "evaluation": {
+    "accuracy": 0.42690763052208835
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/es/can_we_infer_esmt/results.json b/evaluation/xnlimt/es/can_we_infer_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e88590482c46b2a0f9c50c71bbba04503a2c8c8
--- /dev/null
+++ b/evaluation/xnlimt/es/can_we_infer_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "can we infer_esmt",
+  "evaluation": {
+    "accuracy": 0.3895582329317269
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/es/guaranteed_possible_impossible_esmt/results.json b/evaluation/xnlimt/es/guaranteed_possible_impossible_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c45c42fc502a194e4734d1f527f10eff0e86e04
--- /dev/null
+++ b/evaluation/xnlimt/es/guaranteed_possible_impossible_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "guaranteed/possible/impossible_esmt",
+  "evaluation": {
+    "accuracy": 0.3477911646586345
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/es/justified_in_saying_esmt/results.json b/evaluation/xnlimt/es/justified_in_saying_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ab59224bbf868b322f353466f443457107cc276
--- /dev/null
+++ b/evaluation/xnlimt/es/justified_in_saying_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "justified in saying_esmt",
+  "evaluation": {
+    "accuracy": 0.39799196787148594
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/fr/GPT-3_style_frmt/results.json b/evaluation/xnlimt/fr/GPT-3_style_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d588894667d032fd6e29b6ba7e9bcd4024fcd59c
--- /dev/null
+++ b/evaluation/xnlimt/fr/GPT-3_style_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "GPT-3 style_frmt",
+  "evaluation": {
+    "accuracy": 0.5220883534136547
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/fr/MNLI_crowdsource_frmt/results.json b/evaluation/xnlimt/fr/MNLI_crowdsource_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..dd229693dc3267fffb9ffef13310ce4b41f16259
--- /dev/null
+++ b/evaluation/xnlimt/fr/MNLI_crowdsource_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "MNLI crowdsource_frmt",
+  "evaluation": {
+    "accuracy": 0.3192771084337349
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/fr/can_we_infer_frmt/results.json b/evaluation/xnlimt/fr/can_we_infer_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a40ee36a56db7f5039b164cc5f93fc9bf0d4c10e
--- /dev/null
+++ b/evaluation/xnlimt/fr/can_we_infer_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "can we infer_frmt",
+  "evaluation": {
+    "accuracy": 0.5240963855421686
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/fr/guaranteed_possible_impossible_frmt/results.json b/evaluation/xnlimt/fr/guaranteed_possible_impossible_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d3c99fff4f20ce371d4abf4534e0a3de1fbfd1f0
--- /dev/null
+++ b/evaluation/xnlimt/fr/guaranteed_possible_impossible_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "guaranteed/possible/impossible_frmt",
+  "evaluation": {
+    "accuracy": 0.3819277108433735
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/fr/justified_in_saying_frmt/results.json b/evaluation/xnlimt/fr/justified_in_saying_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3c556932c2368ab5dad3ae7ea52e812891e569b1
--- /dev/null
+++ b/evaluation/xnlimt/fr/justified_in_saying_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "justified in saying_frmt",
+  "evaluation": {
+    "accuracy": 0.472289156626506
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/hi/GPT-3_style_himt/results.json b/evaluation/xnlimt/hi/GPT-3_style_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..64d2695e1222b8e2ef70a6be0f451c25b35dc558
--- /dev/null
+++ b/evaluation/xnlimt/hi/GPT-3_style_himt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "GPT-3 style_himt",
+  "evaluation": {
+    "accuracy": 0.3317269076305221
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/hi/MNLI_crowdsource_himt/results.json b/evaluation/xnlimt/hi/MNLI_crowdsource_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..73ae164a98d33cdf0f9767e574f0b858a4f5c78b
--- /dev/null
+++ b/evaluation/xnlimt/hi/MNLI_crowdsource_himt/results.json
@@ -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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/hi/can_we_infer_himt/results.json b/evaluation/xnlimt/hi/can_we_infer_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d20367db7addbcc1a429ed512438bca992387d0d
--- /dev/null
+++ b/evaluation/xnlimt/hi/can_we_infer_himt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "can we infer_himt",
+  "evaluation": {
+    "accuracy": 0.35943775100401604
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/hi/guaranteed_possible_impossible_himt/results.json b/evaluation/xnlimt/hi/guaranteed_possible_impossible_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..34b08178a14f9f6812d8c27edcbeba951547d81e
--- /dev/null
+++ b/evaluation/xnlimt/hi/guaranteed_possible_impossible_himt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "guaranteed/possible/impossible_himt",
+  "evaluation": {
+    "accuracy": 0.3449799196787149
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/hi/justified_in_saying_himt/results.json b/evaluation/xnlimt/hi/justified_in_saying_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1aaa5f37e224219682c46b7e450a46500addd5f4
--- /dev/null
+++ b/evaluation/xnlimt/hi/justified_in_saying_himt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "justified in saying_himt",
+  "evaluation": {
+    "accuracy": 0.3654618473895582
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/merged.csv b/evaluation/xnlimt/merged.csv
new file mode 100644
index 0000000000000000000000000000000000000000..f0ea0a8121ebbea29225f2ac4c3308b146e7b28c
--- /dev/null
+++ b/evaluation/xnlimt/merged.csv
@@ -0,0 +1,50 @@
+dataset,prompt,metric,value
+xnli_ar,GPT-3 style_armt,accuracy,0.3333333333333333
+xnli_ar,MNLI crowdsource_armt,accuracy,0.4855421686746988
+xnli_ar,can we infer_armt,accuracy,0.3413654618473896
+xnli_ar,guaranteed/possible/impossible_armt,accuracy,0.35542168674698793
+xnli_ar,justified in saying_armt,accuracy,0.3465863453815261
+xnli_ar,median,accuracy,0.3465863453815261
+xnli_es,GPT-3 style_esmt,accuracy,0.5385542168674698
+xnli_es,MNLI crowdsource_esmt,accuracy,0.42690763052208835
+xnli_es,can we infer_esmt,accuracy,0.3895582329317269
+xnli_es,guaranteed/possible/impossible_esmt,accuracy,0.3477911646586345
+xnli_es,justified in saying_esmt,accuracy,0.39799196787148594
+xnli_es,median,accuracy,0.39799196787148594
+xnli_fr,GPT-3 style_frmt,accuracy,0.5220883534136547
+xnli_fr,MNLI crowdsource_frmt,accuracy,0.3192771084337349
+xnli_fr,can we infer_frmt,accuracy,0.5240963855421686
+xnli_fr,guaranteed/possible/impossible_frmt,accuracy,0.3819277108433735
+xnli_fr,justified in saying_frmt,accuracy,0.472289156626506
+xnli_fr,median,accuracy,0.472289156626506
+xnli_hi,GPT-3 style_himt,accuracy,0.3317269076305221
+xnli_hi,MNLI crowdsource_himt,accuracy,0.3333333333333333
+xnli_hi,can we infer_himt,accuracy,0.35943775100401604
+xnli_hi,guaranteed/possible/impossible_himt,accuracy,0.3449799196787149
+xnli_hi,justified in saying_himt,accuracy,0.3654618473895582
+xnli_hi,median,accuracy,0.3449799196787149
+xnli_sw,GPT-3 style_swmt,accuracy,0.334136546184739
+xnli_sw,MNLI crowdsource_swmt,accuracy,0.3333333333333333
+xnli_sw,can we infer_swmt,accuracy,0.3337349397590361
+xnli_sw,guaranteed/possible/impossible_swmt,accuracy,0.3261044176706827
+xnli_sw,justified in saying_swmt,accuracy,0.334136546184739
+xnli_sw,median,accuracy,0.3337349397590361
+xnli_ur,GPT-3 style_urmt,accuracy,0.3377510040160643
+xnli_ur,MNLI crowdsource_urmt,accuracy,0.3337349397590361
+xnli_ur,can we infer_urmt,accuracy,0.3333333333333333
+xnli_ur,guaranteed/possible/impossible_urmt,accuracy,0.3333333333333333
+xnli_ur,justified in saying_urmt,accuracy,0.3337349397590361
+xnli_ur,median,accuracy,0.3337349397590361
+xnli_vi,GPT-3 style_vimt,accuracy,0.3333333333333333
+xnli_vi,MNLI crowdsource_vimt,accuracy,0.3887550200803213
+xnli_vi,can we infer_vimt,accuracy,0.3333333333333333
+xnli_vi,guaranteed/possible/impossible_vimt,accuracy,0.3321285140562249
+xnli_vi,justified in saying_vimt,accuracy,0.3333333333333333
+xnli_vi,median,accuracy,0.3333333333333333
+xnli_zh,GPT-3 style_zhmt,accuracy,0.4634538152610442
+xnli_zh,MNLI crowdsource_zhmt,accuracy,0.3345381526104418
+xnli_zh,can we infer_zhmt,accuracy,0.4891566265060241
+xnli_zh,guaranteed/possible/impossible_zhmt,accuracy,0.3393574297188755
+xnli_zh,justified in saying_zhmt,accuracy,0.48032128514056227
+xnli_zh,median,accuracy,0.4634538152610442
+multiple,average,multiple,0.3782630522088353
diff --git a/evaluation/xnlimt/merged.json b/evaluation/xnlimt/merged.json
new file mode 100644
index 0000000000000000000000000000000000000000..12399b33f08c836ec1b9d142e3e493fe304cdf20
--- /dev/null
+++ b/evaluation/xnlimt/merged.json
@@ -0,0 +1 @@
+{"xnli_ar": {"GPT-3 style_armt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "GPT-3 style_armt"}, "MNLI crowdsource_armt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4855421686746988}, "template_name": "MNLI crowdsource_armt"}, "can we infer_armt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3413654618473896}, "template_name": "can we infer_armt"}, "guaranteed/possible/impossible_armt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.35542168674698793}, "template_name": "guaranteed/possible/impossible_armt"}, "justified in saying_armt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3465863453815261}, "template_name": "justified in saying_armt"}}, "xnli_es": {"GPT-3 style_esmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5385542168674698}, "template_name": "GPT-3 style_esmt"}, "MNLI crowdsource_esmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.42690763052208835}, "template_name": "MNLI crowdsource_esmt"}, "can we infer_esmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3895582329317269}, "template_name": "can we infer_esmt"}, "guaranteed/possible/impossible_esmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3477911646586345}, "template_name": "guaranteed/possible/impossible_esmt"}, "justified in saying_esmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.39799196787148594}, "template_name": "justified in saying_esmt"}}, "xnli_fr": {"GPT-3 style_frmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5220883534136547}, "template_name": "GPT-3 style_frmt"}, "MNLI crowdsource_frmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3192771084337349}, "template_name": "MNLI crowdsource_frmt"}, "can we infer_frmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5240963855421686}, "template_name": "can we infer_frmt"}, "guaranteed/possible/impossible_frmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3819277108433735}, "template_name": "guaranteed/possible/impossible_frmt"}, "justified in saying_frmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.472289156626506}, "template_name": "justified in saying_frmt"}}, "xnli_hi": {"GPT-3 style_himt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3317269076305221}, "template_name": "GPT-3 style_himt"}, "MNLI crowdsource_himt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "MNLI crowdsource_himt"}, "can we infer_himt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.35943775100401604}, "template_name": "can we infer_himt"}, "guaranteed/possible/impossible_himt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3449799196787149}, "template_name": "guaranteed/possible/impossible_himt"}, "justified in saying_himt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3654618473895582}, "template_name": "justified in saying_himt"}}, "xnli_sw": {"GPT-3 style_swmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.334136546184739}, "template_name": "GPT-3 style_swmt"}, "MNLI crowdsource_swmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "MNLI crowdsource_swmt"}, "can we infer_swmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3337349397590361}, "template_name": "can we infer_swmt"}, "guaranteed/possible/impossible_swmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3261044176706827}, "template_name": "guaranteed/possible/impossible_swmt"}, "justified in saying_swmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.334136546184739}, "template_name": "justified in saying_swmt"}}, "xnli_ur": {"GPT-3 style_urmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3377510040160643}, "template_name": "GPT-3 style_urmt"}, "MNLI crowdsource_urmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3337349397590361}, "template_name": "MNLI crowdsource_urmt"}, "can we infer_urmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "can we infer_urmt"}, "guaranteed/possible/impossible_urmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "guaranteed/possible/impossible_urmt"}, "justified in saying_urmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3337349397590361}, "template_name": "justified in saying_urmt"}}, "xnli_vi": {"GPT-3 style_vimt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "GPT-3 style_vimt"}, "MNLI crowdsource_vimt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3887550200803213}, "template_name": "MNLI crowdsource_vimt"}, "can we infer_vimt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "can we infer_vimt"}, "guaranteed/possible/impossible_vimt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3321285140562249}, "template_name": "guaranteed/possible/impossible_vimt"}, "justified in saying_vimt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3333333333333333}, "template_name": "justified in saying_vimt"}}, "xnli_zh": {"GPT-3 style_zhmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4634538152610442}, "template_name": "GPT-3 style_zhmt"}, "MNLI crowdsource_zhmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3345381526104418}, "template_name": "MNLI crowdsource_zhmt"}, "can we infer_zhmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4891566265060241}, "template_name": "can we infer_zhmt"}, "guaranteed/possible/impossible_zhmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3393574297188755}, "template_name": "guaranteed/possible/impossible_zhmt"}, "justified in saying_zhmt": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.48032128514056227}, "template_name": "justified in saying_zhmt"}}}
\ No newline at end of file
diff --git a/evaluation/xnlimt/sw/GPT-3_style_swmt/results.json b/evaluation/xnlimt/sw/GPT-3_style_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..5242f09c1270f3f83829ac98159f36869eb2490d
--- /dev/null
+++ b/evaluation/xnlimt/sw/GPT-3_style_swmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "GPT-3 style_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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/sw/MNLI_crowdsource_swmt/results.json b/evaluation/xnlimt/sw/MNLI_crowdsource_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..86775e878263814a2b8403e981105a19842c035d
--- /dev/null
+++ b/evaluation/xnlimt/sw/MNLI_crowdsource_swmt/results.json
@@ -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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/sw/can_we_infer_swmt/results.json b/evaluation/xnlimt/sw/can_we_infer_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b2fb84df09912986eea008200711b1d278d5042
--- /dev/null
+++ b/evaluation/xnlimt/sw/can_we_infer_swmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "can we infer_swmt",
+  "evaluation": {
+    "accuracy": 0.3337349397590361
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/sw/guaranteed_possible_impossible_swmt/results.json b/evaluation/xnlimt/sw/guaranteed_possible_impossible_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..219047a34626d95888a5b5f9099345086a00fa27
--- /dev/null
+++ b/evaluation/xnlimt/sw/guaranteed_possible_impossible_swmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "guaranteed/possible/impossible_swmt",
+  "evaluation": {
+    "accuracy": 0.3261044176706827
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/sw/justified_in_saying_swmt/results.json b/evaluation/xnlimt/sw/justified_in_saying_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d506bf07cfb0240b57bcac472cae7b35158da95b
--- /dev/null
+++ b/evaluation/xnlimt/sw/justified_in_saying_swmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "justified in saying_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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ur/GPT-3_style_urmt/results.json b/evaluation/xnlimt/ur/GPT-3_style_urmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..28cce4c610b89834c6313c33d9972021040e95a4
--- /dev/null
+++ b/evaluation/xnlimt/ur/GPT-3_style_urmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "GPT-3 style_urmt",
+  "evaluation": {
+    "accuracy": 0.3377510040160643
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ur/MNLI_crowdsource_urmt/results.json b/evaluation/xnlimt/ur/MNLI_crowdsource_urmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8f961654aeec2e7a890b280917a887603c772bc9
--- /dev/null
+++ b/evaluation/xnlimt/ur/MNLI_crowdsource_urmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "MNLI crowdsource_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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ur/can_we_infer_urmt/results.json b/evaluation/xnlimt/ur/can_we_infer_urmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..99675a52d98c51024e91224b14cf355c1cae1161
--- /dev/null
+++ b/evaluation/xnlimt/ur/can_we_infer_urmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "can we infer_urmt",
+  "evaluation": {
+    "accuracy": 0.3333333333333333
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ur/guaranteed_possible_impossible_urmt/results.json b/evaluation/xnlimt/ur/guaranteed_possible_impossible_urmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0e5a632d847e31bf188b7a6924e8385f1af5dbd0
--- /dev/null
+++ b/evaluation/xnlimt/ur/guaranteed_possible_impossible_urmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "guaranteed/possible/impossible_urmt",
+  "evaluation": {
+    "accuracy": 0.3333333333333333
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/ur/justified_in_saying_urmt/results.json b/evaluation/xnlimt/ur/justified_in_saying_urmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..12d7fd8973ceeba90f0b3973829fd6297a4bf5e4
--- /dev/null
+++ b/evaluation/xnlimt/ur/justified_in_saying_urmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "justified in saying_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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/vi/GPT-3_style_vimt/results.json b/evaluation/xnlimt/vi/GPT-3_style_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..dfce0056ada3bdb2448acc5af08e6d58192bd49a
--- /dev/null
+++ b/evaluation/xnlimt/vi/GPT-3_style_vimt/results.json
@@ -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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/vi/MNLI_crowdsource_vimt/results.json b/evaluation/xnlimt/vi/MNLI_crowdsource_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b68db45df1b97750e4ca83fbcebe1c2ba51204b0
--- /dev/null
+++ b/evaluation/xnlimt/vi/MNLI_crowdsource_vimt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "MNLI crowdsource_vimt",
+  "evaluation": {
+    "accuracy": 0.3887550200803213
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/vi/can_we_infer_vimt/results.json b/evaluation/xnlimt/vi/can_we_infer_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..399f1e12e2e5e1a960f554afebda1db41b5e54cf
--- /dev/null
+++ b/evaluation/xnlimt/vi/can_we_infer_vimt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "can we infer_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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/vi/guaranteed_possible_impossible_vimt/results.json b/evaluation/xnlimt/vi/guaranteed_possible_impossible_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8555f1f55886cabfadd5a156a2e887a5606cca06
--- /dev/null
+++ b/evaluation/xnlimt/vi/guaranteed_possible_impossible_vimt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "guaranteed/possible/impossible_vimt",
+  "evaluation": {
+    "accuracy": 0.3321285140562249
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/vi/justified_in_saying_vimt/results.json b/evaluation/xnlimt/vi/justified_in_saying_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e1d3100447b06c579e62865f80bac3260b900fb0
--- /dev/null
+++ b/evaluation/xnlimt/vi/justified_in_saying_vimt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "justified in saying_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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/zh/GPT-3_style_zhmt/results.json b/evaluation/xnlimt/zh/GPT-3_style_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a9e431c729c8176051311252cb6df006feb918b8
--- /dev/null
+++ b/evaluation/xnlimt/zh/GPT-3_style_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "GPT-3 style_zhmt",
+  "evaluation": {
+    "accuracy": 0.4634538152610442
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/zh/MNLI_crowdsource_zhmt/results.json b/evaluation/xnlimt/zh/MNLI_crowdsource_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..c8548b52fac45ccbd571c2f3c5ddf1de6d342e04
--- /dev/null
+++ b/evaluation/xnlimt/zh/MNLI_crowdsource_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "MNLI crowdsource_zhmt",
+  "evaluation": {
+    "accuracy": 0.3345381526104418
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/zh/can_we_infer_zhmt/results.json b/evaluation/xnlimt/zh/can_we_infer_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..469b0a1ae48467123059fdd73ca7a3aa38150589
--- /dev/null
+++ b/evaluation/xnlimt/zh/can_we_infer_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "can we infer_zhmt",
+  "evaluation": {
+    "accuracy": 0.4891566265060241
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/zh/guaranteed_possible_impossible_zhmt/results.json b/evaluation/xnlimt/zh/guaranteed_possible_impossible_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..85c0552653669d8b485159da28454b6b60e4cb91
--- /dev/null
+++ b/evaluation/xnlimt/zh/guaranteed_possible_impossible_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "guaranteed/possible/impossible_zhmt",
+  "evaluation": {
+    "accuracy": 0.3393574297188755
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation/xnlimt/zh/justified_in_saying_zhmt/results.json b/evaluation/xnlimt/zh/justified_in_saying_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..f90bef80a421b6c9c19603659c8c32d751877973
--- /dev/null
+++ b/evaluation/xnlimt/zh/justified_in_saying_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "justified in saying_zhmt",
+  "evaluation": {
+    "accuracy": 0.48032128514056227
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/ar/Answer_Given_options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/ar/Answer_Given_options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ee687fff025db235d1a0c5e3e7e4d5ea37b70fa5
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/ar/Answer_Given_options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Answer Given options",
+  "evaluation": {
+    "accuracy": 0.6896095301125083
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/ar/Choose_Story_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/ar/Choose_Story_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..09913ae524e676cae982ea0bf63fafafde804167
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/ar/Choose_Story_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Choose Story Ending",
+  "evaluation": {
+    "accuracy": 0.8378557246856386
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/ar/Generate_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/ar/Generate_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..357752fcef715dd0e9bb3e53f19541a6e4677762
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/ar/Generate_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Generate Ending",
+  "evaluation": {
+    "accuracy": 0.5956320317670417
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/ar/Novel_Correct_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/ar/Novel_Correct_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8efe6a29767e7402a35636542d83d0ca0dcc99aa
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/ar/Novel_Correct_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Novel Correct Ending",
+  "evaluation": {
+    "accuracy": 0.8213103904698875
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/ar/Story_Continuation_and_Options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/ar/Story_Continuation_and_Options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ab27c96b2527d678ce13c15cfdbde2a31adebef5
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/ar/Story_Continuation_and_Options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Story Continuation and Options",
+  "evaluation": {
+    "accuracy": 0.8219722038385175
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/es/Answer_Given_options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/es/Answer_Given_options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e752bca41c96f3e30dfe70484266525ef56f78d6
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/es/Answer_Given_options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Answer Given options",
+  "evaluation": {
+    "accuracy": 0.7683653209794837
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/es/Choose_Story_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/es/Choose_Story_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..fca6014811f0daaa82752ddaaa993561ce108c44
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/es/Choose_Story_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Choose Story Ending",
+  "evaluation": {
+    "accuracy": 0.886168100595632
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/es/Generate_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/es/Generate_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..f3777d75e7650424308f106fd6d3a5df7f787d90
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/es/Generate_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Generate Ending",
+  "evaluation": {
+    "accuracy": 0.6724023825281271
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/es/Novel_Correct_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/es/Novel_Correct_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..53b09d0e87a347427a568a9a833c4fd900355f1b
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/es/Novel_Correct_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Novel Correct Ending",
+  "evaluation": {
+    "accuracy": 0.8676373262739907
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/es/Story_Continuation_and_Options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/es/Story_Continuation_and_Options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..f9518581ecda9f5cf8b5b17defb0740e70dd9640
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/es/Story_Continuation_and_Options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Story Continuation and Options",
+  "evaluation": {
+    "accuracy": 0.8769027134348114
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/eu/Answer_Given_options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/eu/Answer_Given_options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a6bf025a2d3ea4baeb360ae84d4b8b69e15a024c
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/eu/Answer_Given_options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Answer Given options",
+  "evaluation": {
+    "accuracy": 0.6082064857710126
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/eu/Choose_Story_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/eu/Choose_Story_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0632573e4226ca0469eb8baa954d9a633b95c867
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/eu/Choose_Story_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Choose Story Ending",
+  "evaluation": {
+    "accuracy": 0.7266710787557908
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/eu/Generate_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/eu/Generate_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2c7e2be9cbc4fc0796d38b369755eb1270b944d8
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/eu/Generate_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Generate Ending",
+  "evaluation": {
+    "accuracy": 0.5552614162806089
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/eu/Novel_Correct_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/eu/Novel_Correct_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..386f9bca23da50f8812d5a65c99a454e3bfd3a84
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/eu/Novel_Correct_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Novel Correct Ending",
+  "evaluation": {
+    "accuracy": 0.700198544010589
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/eu/Story_Continuation_and_Options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/eu/Story_Continuation_and_Options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a706ebc93e3866716af502ede850131876542e30
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/eu/Story_Continuation_and_Options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Story Continuation and Options",
+  "evaluation": {
+    "accuracy": 0.7107875579086698
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/hi/Answer_Given_options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/hi/Answer_Given_options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..36c5bb8484b6d505c4afb870897acae7619e6958
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/hi/Answer_Given_options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Answer Given options",
+  "evaluation": {
+    "accuracy": 0.6366644606221046
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/hi/Choose_Story_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/hi/Choose_Story_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..5ed2515ae17faae85ed0bc6a96c0d985ed4bddb5
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/hi/Choose_Story_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Choose Story Ending",
+  "evaluation": {
+    "accuracy": 0.7882197220383852
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/hi/Generate_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/hi/Generate_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d807e21b29de04446f4cfe611ab4a9f1c1aac6f3
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/hi/Generate_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Generate Ending",
+  "evaluation": {
+    "accuracy": 0.5982792852415619
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/hi/Novel_Correct_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/hi/Novel_Correct_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b43a6d1cd13493fd76b1f429cf6b268d8fb3eae9
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/hi/Novel_Correct_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Novel Correct Ending",
+  "evaluation": {
+    "accuracy": 0.7485109199205824
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/hi/Story_Continuation_and_Options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/hi/Story_Continuation_and_Options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..5d3b8d42e9a6c3e33a1933b8157ddcb0e1e3d5b3
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/hi/Story_Continuation_and_Options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Story Continuation and Options",
+  "evaluation": {
+    "accuracy": 0.7683653209794837
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/id/Answer_Given_options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/id/Answer_Given_options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..42973d245d09708e0cb10381312f81c1a009fd37
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/id/Answer_Given_options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Answer Given options",
+  "evaluation": {
+    "accuracy": 0.7385837193911317
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/id/Choose_Story_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/id/Choose_Story_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2083ffe7418714a31fc494282fb2a5d5d913dd16
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/id/Choose_Story_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Choose Story Ending",
+  "evaluation": {
+    "accuracy": 0.8332230311052283
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/id/Generate_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/id/Generate_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0553cf8d1be7424ae76ed745bacef1655f0af56c
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/id/Generate_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Generate Ending",
+  "evaluation": {
+    "accuracy": 0.6293845135671741
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/id/Novel_Correct_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/id/Novel_Correct_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6884f7ddd6987951ac927728b6032991cbdd40d1
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/id/Novel_Correct_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Novel Correct Ending",
+  "evaluation": {
+    "accuracy": 0.7816015883520847
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/id/Story_Continuation_and_Options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/id/Story_Continuation_and_Options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..623362eb25af762a34a6adced1ec42cb97ec4b41
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/id/Story_Continuation_and_Options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Story Continuation and Options",
+  "evaluation": {
+    "accuracy": 0.8226340172071476
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/zh/Answer_Given_options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/zh/Answer_Given_options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..9197b8cb23453249367d3b46eff4c0c6ec7f5291
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/zh/Answer_Given_options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Answer Given options",
+  "evaluation": {
+    "accuracy": 0.7498345466578424
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/zh/Choose_Story_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/zh/Choose_Story_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..672e6984b4f37403144ee2bfabd0f09d2c190db3
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/zh/Choose_Story_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Choose Story Ending",
+  "evaluation": {
+    "accuracy": 0.8583719391131701
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/zh/Generate_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/zh/Generate_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..21438162a277b6d07ae6068a23aca83bc6388c96
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/zh/Generate_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Generate Ending",
+  "evaluation": {
+    "accuracy": 0.6227663798808736
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/zh/Novel_Correct_Ending/results.json b/evaluation_l1/Muennighoff_xstory_cloze/zh/Novel_Correct_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..f7d3e0fa656be6fcab419b168f75281f7934d963
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/zh/Novel_Correct_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Novel Correct Ending",
+  "evaluation": {
+    "accuracy": 0.8405029781601588
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xstory_cloze/zh/Story_Continuation_and_Options/results.json b/evaluation_l1/Muennighoff_xstory_cloze/zh/Story_Continuation_and_Options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8e57f15ca55b9773a5f03692f5e8a03d17676d04
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xstory_cloze/zh/Story_Continuation_and_Options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Story Continuation and Options",
+  "evaluation": {
+    "accuracy": 0.8385175380542687
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/en/Replace/results.json b/evaluation_l1/Muennighoff_xwinograd/en/Replace/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7cca016bfa305e04743256d35dde6e1dda3b1b4f
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/en/Replace/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "en",
+  "template_name": "Replace",
+  "evaluation": {
+    "accuracy": 0.6576344086021505
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/en/True_or_False/results.json b/evaluation_l1/Muennighoff_xwinograd/en/True_or_False/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..c438a5bc83423e5519bd90ba7ecff0dc88d0cc5e
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/en/True_or_False/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "en",
+  "template_name": "True or False",
+  "evaluation": {
+    "accuracy": 0.5187096774193548
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/en/does_underscore_refer_to/results.json b/evaluation_l1/Muennighoff_xwinograd/en/does_underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0406ac2b9a66283e0e273eb8136dfd7014e870df
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/en/does_underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "en",
+  "template_name": "does underscore refer to",
+  "evaluation": {
+    "accuracy": 0.5931182795698925
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/en/stand_for/results.json b/evaluation_l1/Muennighoff_xwinograd/en/stand_for/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d55f5f2091fe97a7f784f08252944c43ec04ec5
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/en/stand_for/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "en",
+  "template_name": "stand for",
+  "evaluation": {
+    "accuracy": 0.5070967741935484
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/en/underscore_refer_to/results.json b/evaluation_l1/Muennighoff_xwinograd/en/underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..cbbd7c31c32708d98dbbeb064a418844ed5a841f
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/en/underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "en",
+  "template_name": "underscore refer to",
+  "evaluation": {
+    "accuracy": 0.6210752688172043
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/fr/Replace/results.json b/evaluation_l1/Muennighoff_xwinograd/fr/Replace/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a5b58afaeaf436677903e5cebd1e0d95398c2e1
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/fr/Replace/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "Replace",
+  "evaluation": {
+    "accuracy": 0.5180722891566265
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/fr/True_or_False/results.json b/evaluation_l1/Muennighoff_xwinograd/fr/True_or_False/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..5fcca5729178fa86a2db2771c52b208d5300ce29
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/fr/True_or_False/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "True or False",
+  "evaluation": {
+    "accuracy": 0.5301204819277109
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/fr/does_underscore_refer_to/results.json b/evaluation_l1/Muennighoff_xwinograd/fr/does_underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e633ec7e691facaa5d58953c2c1dda2ff8d7244d
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/fr/does_underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "does underscore refer to",
+  "evaluation": {
+    "accuracy": 0.5542168674698795
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/fr/stand_for/results.json b/evaluation_l1/Muennighoff_xwinograd/fr/stand_for/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..84c78ab68f45d54a9bc57fa5bd2db69a8a3eee1c
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/fr/stand_for/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "stand for",
+  "evaluation": {
+    "accuracy": 0.5180722891566265
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/fr/underscore_refer_to/results.json b/evaluation_l1/Muennighoff_xwinograd/fr/underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1c8671c7e91fcfa90f5c712c4cd45bd21bce75b6
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/fr/underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "underscore refer to",
+  "evaluation": {
+    "accuracy": 0.5421686746987951
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/pt/Replace/results.json b/evaluation_l1/Muennighoff_xwinograd/pt/Replace/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..40eb613c0db3e719a7c14f420c93feaca8db71b9
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/pt/Replace/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "Replace",
+  "evaluation": {
+    "accuracy": 0.5741444866920152
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/pt/True_or_False/results.json b/evaluation_l1/Muennighoff_xwinograd/pt/True_or_False/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..56359460ad660a81be9fb14f1b24a16bc423c553
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/pt/True_or_False/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "True or False",
+  "evaluation": {
+    "accuracy": 0.4790874524714829
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/pt/does_underscore_refer_to/results.json b/evaluation_l1/Muennighoff_xwinograd/pt/does_underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..72c0b538422b56ee2f446ae55334d7f16aa1d5a3
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/pt/does_underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "does underscore refer to",
+  "evaluation": {
+    "accuracy": 0.55893536121673
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/pt/stand_for/results.json b/evaluation_l1/Muennighoff_xwinograd/pt/stand_for/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3480298da420f74bcc47be59736b849b7b1d16fd
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/pt/stand_for/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "stand for",
+  "evaluation": {
+    "accuracy": 0.5209125475285171
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/pt/underscore_refer_to/results.json b/evaluation_l1/Muennighoff_xwinograd/pt/underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e80e9293322a6d429f21f5f14f68c9303eafe211
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/pt/underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "underscore refer to",
+  "evaluation": {
+    "accuracy": 0.5437262357414449
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/zh/Replace/results.json b/evaluation_l1/Muennighoff_xwinograd/zh/Replace/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3ca8099b5daa44ef38338971f64730faf00e56bf
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/zh/Replace/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "Replace",
+  "evaluation": {
+    "accuracy": 0.626984126984127
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/zh/True_or_False/results.json b/evaluation_l1/Muennighoff_xwinograd/zh/True_or_False/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..4d7ecc65170e29e452aed31e5668815572903efd
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/zh/True_or_False/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "True or False",
+  "evaluation": {
+    "accuracy": 0.503968253968254
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/zh/does_underscore_refer_to/results.json b/evaluation_l1/Muennighoff_xwinograd/zh/does_underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..fbdd04009638a4026303cb221180e18c7179dc8b
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/zh/does_underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "does underscore refer to",
+  "evaluation": {
+    "accuracy": 0.5436507936507936
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/zh/stand_for/results.json b/evaluation_l1/Muennighoff_xwinograd/zh/stand_for/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6a6da375fe704c9bb9eeac0c8304879e1fc26865
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/zh/stand_for/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "stand for",
+  "evaluation": {
+    "accuracy": 0.49007936507936506
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/Muennighoff_xwinograd/zh/underscore_refer_to/results.json b/evaluation_l1/Muennighoff_xwinograd/zh/underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3e946b644795bdac06366e70df599e5c70ef095d
--- /dev/null
+++ b/evaluation_l1/Muennighoff_xwinograd/zh/underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "underscore refer to",
+  "evaluation": {
+    "accuracy": 0.5535714285714286
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r1/GPT-3_style/results.json b/evaluation_l1/anli/dev_r1/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7d797684c48813fb93ee7fcee75012e9220216bd
--- /dev/null
+++ b/evaluation_l1/anli/dev_r1/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r1",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.426
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r1/MNLI_crowdsource/results.json b/evaluation_l1/anli/dev_r1/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..20374d1d4070ba04732b4c4875762233e3b51afa
--- /dev/null
+++ b/evaluation_l1/anli/dev_r1/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r1",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.402
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r1/can_we_infer/results.json b/evaluation_l1/anli/dev_r1/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6a28787134fdaee4350e53df323e24eb4336c101
--- /dev/null
+++ b/evaluation_l1/anli/dev_r1/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r1",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.401
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r1/guaranteed_possible_impossible/results.json b/evaluation_l1/anli/dev_r1/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b12be3ab4bc1fe7034c070a3cd0a856abcdaff89
--- /dev/null
+++ b/evaluation_l1/anli/dev_r1/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r1",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.314
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r1/justified_in_saying/results.json b/evaluation_l1/anli/dev_r1/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6893460fc25c185b0f863c0771d50a331a924574
--- /dev/null
+++ b/evaluation_l1/anli/dev_r1/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r1",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.387
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r2/GPT-3_style/results.json b/evaluation_l1/anli/dev_r2/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..222b602340107db392e64549acd17451eef3b73a
--- /dev/null
+++ b/evaluation_l1/anli/dev_r2/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r2",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.383
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r2/MNLI_crowdsource/results.json b/evaluation_l1/anli/dev_r2/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..5c7dd6e91ff9d27c32c5412ac78786cf7d2d7c02
--- /dev/null
+++ b/evaluation_l1/anli/dev_r2/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r2",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.374
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r2/can_we_infer/results.json b/evaluation_l1/anli/dev_r2/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..af61a800ad5c00e1df7087f25d7189bcd54ed72e
--- /dev/null
+++ b/evaluation_l1/anli/dev_r2/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r2",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.394
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r2/guaranteed_possible_impossible/results.json b/evaluation_l1/anli/dev_r2/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..05a872ed6fea16bda1aa0c8cdb6ab02e80e036ca
--- /dev/null
+++ b/evaluation_l1/anli/dev_r2/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r2",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.302
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r2/justified_in_saying/results.json b/evaluation_l1/anli/dev_r2/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2672da6c326b4a26b10b029420fc95e9947f5608
--- /dev/null
+++ b/evaluation_l1/anli/dev_r2/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r2",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.376
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r3/GPT-3_style/results.json b/evaluation_l1/anli/dev_r3/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..f151cc8d2e07b9e0b3ef1edf849dd69a8c748873
--- /dev/null
+++ b/evaluation_l1/anli/dev_r3/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r3",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.42
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r3/MNLI_crowdsource/results.json b/evaluation_l1/anli/dev_r3/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..82148467711e65c52e039e9346eaecb5d7f3ee84
--- /dev/null
+++ b/evaluation_l1/anli/dev_r3/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r3",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.4116666666666667
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r3/can_we_infer/results.json b/evaluation_l1/anli/dev_r3/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..21fdd815d985648d60398c318c503b30dc209554
--- /dev/null
+++ b/evaluation_l1/anli/dev_r3/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r3",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.38916666666666666
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r3/guaranteed_possible_impossible/results.json b/evaluation_l1/anli/dev_r3/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8c95797516b28c82b8d3598a8a5dda398e84e5f0
--- /dev/null
+++ b/evaluation_l1/anli/dev_r3/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r3",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.2966666666666667
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/anli/dev_r3/justified_in_saying/results.json b/evaluation_l1/anli/dev_r3/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ec7cea823be814a3a1c396d7e2e6121f6e6cb1ee
--- /dev/null
+++ b/evaluation_l1/anli/dev_r3/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "anli",
+  "dataset_config_name": "dev_r3",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.35833333333333334
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/merged.csv b/evaluation_l1/merged.csv
new file mode 100644
index 0000000000000000000000000000000000000000..7786678ad3e7adbcb02bec034f8fa4362a1bc340
--- /dev/null
+++ b/evaluation_l1/merged.csv
@@ -0,0 +1,194 @@
+dataset,prompt,metric,value
+anli_dev_r1,GPT-3 style,accuracy,0.426
+anli_dev_r1,MNLI crowdsource,accuracy,0.402
+anli_dev_r1,can we infer,accuracy,0.401
+anli_dev_r1,guaranteed/possible/impossible,accuracy,0.314
+anli_dev_r1,justified in saying,accuracy,0.387
+anli_dev_r1,median,accuracy,0.401
+anli_dev_r2,GPT-3 style,accuracy,0.383
+anli_dev_r2,MNLI crowdsource,accuracy,0.374
+anli_dev_r2,can we infer,accuracy,0.394
+anli_dev_r2,guaranteed/possible/impossible,accuracy,0.302
+anli_dev_r2,justified in saying,accuracy,0.376
+anli_dev_r2,median,accuracy,0.376
+anli_dev_r3,GPT-3 style,accuracy,0.42
+anli_dev_r3,MNLI crowdsource,accuracy,0.4116666666666667
+anli_dev_r3,can we infer,accuracy,0.38916666666666666
+anli_dev_r3,guaranteed/possible/impossible,accuracy,0.2966666666666667
+anli_dev_r3,justified in saying,accuracy,0.35833333333333334
+anli_dev_r3,median,accuracy,0.38916666666666666
+story_cloze_2016,Answer Given options,accuracy,0.8524853019775521
+story_cloze_2016,Choose Story Ending,accuracy,0.8957776590058792
+story_cloze_2016,Generate Ending,accuracy,0.709246392303581
+story_cloze_2016,Novel Correct Ending,accuracy,0.8888295029396045
+story_cloze_2016,Story Continuation and Options,accuracy,0.8850881881346874
+story_cloze_2016,median,accuracy,0.8850881881346874
+super_glue_cb,GPT-3 style,accuracy,0.8392857142857143
+super_glue_cb,MNLI crowdsource,accuracy,0.35714285714285715
+super_glue_cb,can we infer,accuracy,0.7857142857142857
+super_glue_cb,guaranteed/possible/impossible,accuracy,0.5535714285714286
+super_glue_cb,justified in saying,accuracy,0.7142857142857143
+super_glue_cb,median,accuracy,0.7142857142857143
+super_glue_copa,"C1 or C2? premise, so/because…",accuracy,0.66
+super_glue_copa,best_option,accuracy,0.77
+super_glue_copa,cause_effect,accuracy,0.8
+super_glue_copa,i_am_hesitating,accuracy,0.81
+super_glue_copa,plausible_alternatives,accuracy,0.84
+super_glue_copa,median,accuracy,0.8
+super_glue_rte,GPT-3 style,accuracy,0.7906137184115524
+super_glue_rte,MNLI crowdsource,accuracy,0.8267148014440433
+super_glue_rte,does it follow that,accuracy,0.7942238267148014
+super_glue_rte,guaranteed true,accuracy,0.776173285198556
+super_glue_rte,should assume,accuracy,0.7617328519855595
+super_glue_rte,median,accuracy,0.7906137184115524
+winogrande_winogrande_xl,Replace,accuracy,0.5588003157063931
+winogrande_winogrande_xl,True or False,accuracy,0.5280189423835833
+winogrande_winogrande_xl,does underscore refer to,accuracy,0.5651144435674822
+winogrande_winogrande_xl,stand for,accuracy,0.5082872928176796
+winogrande_winogrande_xl,underscore refer to,accuracy,0.5651144435674822
+winogrande_winogrande_xl,median,accuracy,0.5588003157063931
+xcopa_id,"C1 or C2? premise, so/because…",accuracy,0.46
+xcopa_id,best_option,accuracy,0.7
+xcopa_id,cause_effect,accuracy,0.73
+xcopa_id,i_am_hesitating,accuracy,0.72
+xcopa_id,plausible_alternatives,accuracy,0.67
+xcopa_id,median,accuracy,0.7
+xcopa_sw,"C1 or C2? premise, so/because…",accuracy,0.6
+xcopa_sw,best_option,accuracy,0.55
+xcopa_sw,cause_effect,accuracy,0.54
+xcopa_sw,i_am_hesitating,accuracy,0.51
+xcopa_sw,plausible_alternatives,accuracy,0.52
+xcopa_sw,median,accuracy,0.54
+xcopa_ta,"C1 or C2? premise, so/because…",accuracy,0.59
+xcopa_ta,best_option,accuracy,0.56
+xcopa_ta,cause_effect,accuracy,0.6
+xcopa_ta,i_am_hesitating,accuracy,0.57
+xcopa_ta,plausible_alternatives,accuracy,0.62
+xcopa_ta,median,accuracy,0.59
+xcopa_vi,"C1 or C2? premise, so/because…",accuracy,0.53
+xcopa_vi,best_option,accuracy,0.72
+xcopa_vi,cause_effect,accuracy,0.72
+xcopa_vi,i_am_hesitating,accuracy,0.7
+xcopa_vi,plausible_alternatives,accuracy,0.71
+xcopa_vi,median,accuracy,0.71
+xcopa_zh,"C1 or C2? premise, so/because…",accuracy,0.67
+xcopa_zh,best_option,accuracy,0.7
+xcopa_zh,cause_effect,accuracy,0.8
+xcopa_zh,i_am_hesitating,accuracy,0.77
+xcopa_zh,plausible_alternatives,accuracy,0.79
+xcopa_zh,median,accuracy,0.77
+xnli_ar,GPT-3 style,accuracy,0.5558232931726907
+xnli_ar,MNLI crowdsource,accuracy,0.42128514056224897
+xnli_ar,can we infer,accuracy,0.5148594377510041
+xnli_ar,guaranteed/possible/impossible,accuracy,0.40562248995983935
+xnli_ar,justified in saying,accuracy,0.4927710843373494
+xnli_ar,median,accuracy,0.4927710843373494
+xnli_en,GPT-3 style,accuracy,0.5891566265060241
+xnli_en,MNLI crowdsource,accuracy,0.42610441767068274
+xnli_en,can we infer,accuracy,0.5662650602409639
+xnli_en,guaranteed/possible/impossible,accuracy,0.4614457831325301
+xnli_en,justified in saying,accuracy,0.5437751004016064
+xnli_en,median,accuracy,0.5437751004016064
+xnli_es,GPT-3 style,accuracy,0.5734939759036145
+xnli_es,MNLI crowdsource,accuracy,0.40923694779116465
+xnli_es,can we infer,accuracy,0.5148594377510041
+xnli_es,guaranteed/possible/impossible,accuracy,0.43132530120481927
+xnli_es,justified in saying,accuracy,0.4610441767068273
+xnli_es,median,accuracy,0.4610441767068273
+xnli_fr,GPT-3 style,accuracy,0.5666666666666667
+xnli_fr,MNLI crowdsource,accuracy,0.42208835341365464
+xnli_fr,can we infer,accuracy,0.5385542168674698
+xnli_fr,guaranteed/possible/impossible,accuracy,0.39076305220883534
+xnli_fr,justified in saying,accuracy,0.5100401606425703
+xnli_fr,median,accuracy,0.5100401606425703
+xnli_hi,GPT-3 style,accuracy,0.5345381526104418
+xnli_hi,MNLI crowdsource,accuracy,0.41124497991967873
+xnli_hi,can we infer,accuracy,0.4751004016064257
+xnli_hi,guaranteed/possible/impossible,accuracy,0.40923694779116465
+xnli_hi,justified in saying,accuracy,0.4469879518072289
+xnli_hi,median,accuracy,0.4469879518072289
+xnli_sw,GPT-3 style,accuracy,0.4827309236947791
+xnli_sw,MNLI crowdsource,accuracy,0.40562248995983935
+xnli_sw,can we infer,accuracy,0.44497991967871486
+xnli_sw,guaranteed/possible/impossible,accuracy,0.42289156626506025
+xnli_sw,justified in saying,accuracy,0.41124497991967873
+xnli_sw,median,accuracy,0.42289156626506025
+xnli_ur,GPT-3 style,accuracy,0.4947791164658635
+xnli_ur,MNLI crowdsource,accuracy,0.39759036144578314
+xnli_ur,can we infer,accuracy,0.4502008032128514
+xnli_ur,guaranteed/possible/impossible,accuracy,0.39036144578313253
+xnli_ur,justified in saying,accuracy,0.40843373493975904
+xnli_ur,median,accuracy,0.40843373493975904
+xnli_vi,GPT-3 style,accuracy,0.5449799196787148
+xnli_vi,MNLI crowdsource,accuracy,0.40401606425702813
+xnli_vi,can we infer,accuracy,0.5
+xnli_vi,guaranteed/possible/impossible,accuracy,0.44779116465863456
+xnli_vi,justified in saying,accuracy,0.4650602409638554
+xnli_vi,median,accuracy,0.4650602409638554
+xnli_zh,GPT-3 style,accuracy,0.5429718875502008
+xnli_zh,MNLI crowdsource,accuracy,0.3891566265060241
+xnli_zh,can we infer,accuracy,0.5032128514056224
+xnli_zh,guaranteed/possible/impossible,accuracy,0.38072289156626504
+xnli_zh,justified in saying,accuracy,0.4706827309236948
+xnli_zh,median,accuracy,0.4706827309236948
+xstory_cloze_ar,Answer Given options,accuracy,0.6896095301125083
+xstory_cloze_ar,Choose Story Ending,accuracy,0.8378557246856386
+xstory_cloze_ar,Generate Ending,accuracy,0.5956320317670417
+xstory_cloze_ar,Novel Correct Ending,accuracy,0.8213103904698875
+xstory_cloze_ar,Story Continuation and Options,accuracy,0.8219722038385175
+xstory_cloze_ar,median,accuracy,0.8213103904698875
+xstory_cloze_es,Answer Given options,accuracy,0.7683653209794837
+xstory_cloze_es,Choose Story Ending,accuracy,0.886168100595632
+xstory_cloze_es,Generate Ending,accuracy,0.6724023825281271
+xstory_cloze_es,Novel Correct Ending,accuracy,0.8676373262739907
+xstory_cloze_es,Story Continuation and Options,accuracy,0.8769027134348114
+xstory_cloze_es,median,accuracy,0.8676373262739907
+xstory_cloze_eu,Answer Given options,accuracy,0.6082064857710126
+xstory_cloze_eu,Choose Story Ending,accuracy,0.7266710787557908
+xstory_cloze_eu,Generate Ending,accuracy,0.5552614162806089
+xstory_cloze_eu,Novel Correct Ending,accuracy,0.700198544010589
+xstory_cloze_eu,Story Continuation and Options,accuracy,0.7107875579086698
+xstory_cloze_eu,median,accuracy,0.700198544010589
+xstory_cloze_hi,Answer Given options,accuracy,0.6366644606221046
+xstory_cloze_hi,Choose Story Ending,accuracy,0.7882197220383852
+xstory_cloze_hi,Generate Ending,accuracy,0.5982792852415619
+xstory_cloze_hi,Novel Correct Ending,accuracy,0.7485109199205824
+xstory_cloze_hi,Story Continuation and Options,accuracy,0.7683653209794837
+xstory_cloze_hi,median,accuracy,0.7485109199205824
+xstory_cloze_id,Answer Given options,accuracy,0.7385837193911317
+xstory_cloze_id,Choose Story Ending,accuracy,0.8332230311052283
+xstory_cloze_id,Generate Ending,accuracy,0.6293845135671741
+xstory_cloze_id,Novel Correct Ending,accuracy,0.7816015883520847
+xstory_cloze_id,Story Continuation and Options,accuracy,0.8226340172071476
+xstory_cloze_id,median,accuracy,0.7816015883520847
+xstory_cloze_zh,Answer Given options,accuracy,0.7498345466578424
+xstory_cloze_zh,Choose Story Ending,accuracy,0.8583719391131701
+xstory_cloze_zh,Generate Ending,accuracy,0.6227663798808736
+xstory_cloze_zh,Novel Correct Ending,accuracy,0.8405029781601588
+xstory_cloze_zh,Story Continuation and Options,accuracy,0.8385175380542687
+xstory_cloze_zh,median,accuracy,0.8385175380542687
+xwinograd_en,Replace,accuracy,0.6576344086021505
+xwinograd_en,True or False,accuracy,0.5187096774193548
+xwinograd_en,does underscore refer to,accuracy,0.5931182795698925
+xwinograd_en,stand for,accuracy,0.5070967741935484
+xwinograd_en,underscore refer to,accuracy,0.6210752688172043
+xwinograd_en,median,accuracy,0.5931182795698925
+xwinograd_fr,Replace,accuracy,0.5180722891566265
+xwinograd_fr,True or False,accuracy,0.5301204819277109
+xwinograd_fr,does underscore refer to,accuracy,0.5542168674698795
+xwinograd_fr,stand for,accuracy,0.5180722891566265
+xwinograd_fr,underscore refer to,accuracy,0.5421686746987951
+xwinograd_fr,median,accuracy,0.5301204819277109
+xwinograd_pt,Replace,accuracy,0.5741444866920152
+xwinograd_pt,True or False,accuracy,0.4790874524714829
+xwinograd_pt,does underscore refer to,accuracy,0.55893536121673
+xwinograd_pt,stand for,accuracy,0.5209125475285171
+xwinograd_pt,underscore refer to,accuracy,0.5437262357414449
+xwinograd_pt,median,accuracy,0.5437262357414449
+xwinograd_zh,Replace,accuracy,0.626984126984127
+xwinograd_zh,True or False,accuracy,0.503968253968254
+xwinograd_zh,does underscore refer to,accuracy,0.5436507936507936
+xwinograd_zh,stand for,accuracy,0.49007936507936506
+xwinograd_zh,underscore refer to,accuracy,0.5535714285714286
+xwinograd_zh,median,accuracy,0.5436507936507936
+multiple,average,multiple,0.6067197952551315
diff --git a/evaluation_l1/merged.json b/evaluation_l1/merged.json
new file mode 100644
index 0000000000000000000000000000000000000000..c58141c131fb8e3ab2c2f7a6c9c1f0a021862123
--- /dev/null
+++ b/evaluation_l1/merged.json
@@ -0,0 +1 @@
+{"Muennighoff/xstory_cloze_ar": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6896095301125083}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8378557246856386}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.5956320317670417}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8213103904698875}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8219722038385175}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_es": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7683653209794837}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.886168100595632}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6724023825281271}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8676373262739907}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8769027134348114}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_eu": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6082064857710126}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7266710787557908}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.5552614162806089}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.700198544010589}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7107875579086698}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_hi": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6366644606221046}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7882197220383852}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.5982792852415619}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7485109199205824}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7683653209794837}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_id": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7385837193911317}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8332230311052283}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6293845135671741}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7816015883520847}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8226340172071476}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xstory_cloze_zh": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7498345466578424}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8583719391131701}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6227663798808736}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8405029781601588}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8385175380542687}, "template_name": "Story Continuation and Options"}}, "Muennighoff/xwinograd_en": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6576344086021505}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5187096774193548}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5931182795698925}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5070967741935484}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6210752688172043}, "template_name": "underscore refer to"}}, "Muennighoff/xwinograd_fr": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5180722891566265}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5301204819277109}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5542168674698795}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5180722891566265}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5421686746987951}, "template_name": "underscore refer to"}}, "Muennighoff/xwinograd_pt": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5741444866920152}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.4790874524714829}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.55893536121673}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5209125475285171}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5437262357414449}, "template_name": "underscore refer to"}}, "Muennighoff/xwinograd_zh": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.626984126984127}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.503968253968254}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5436507936507936}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.49007936507936506}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='en', template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5535714285714286}, "template_name": "underscore refer to"}}, "anli_dev_r1": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.426}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.402}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.401}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.314}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r1', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r1', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r1", "dataset_name": "anli", "evaluation": {"accuracy": 0.387}, "template_name": "justified in saying"}}, "anli_dev_r2": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.383}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.374}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.394}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.302}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r2', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r2', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r2", "dataset_name": "anli", "evaluation": {"accuracy": 0.376}, "template_name": "justified in saying"}}, "anli_dev_r3": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.42}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.4116666666666667}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.38916666666666666}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.2966666666666667}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='dev_r3', dataset_name='anli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='dev_r3', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "dev_r3", "dataset_name": "anli", "evaluation": {"accuracy": 0.35833333333333334}, "template_name": "justified in saying"}}, "story_cloze_2016": {"Answer Given options": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.8524853019775521}, "template_name": "Answer Given options"}, "Choose Story Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.8957776590058792}, "template_name": "Choose Story Ending"}, "Generate Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.709246392303581}, "template_name": "Generate Ending"}, "Novel Correct Ending": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.8888295029396045}, "template_name": "Novel Correct Ending"}, "Story Continuation and Options": {"arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "2016", "dataset_name": "story_cloze", "evaluation": {"accuracy": 0.8850881881346874}, "template_name": "Story Continuation and Options"}}, "super_glue_cb": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.8392857142857143}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.35714285714285715}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.7857142857142857}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.5535714285714286}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "cb", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.7142857142857143}, "template_name": "justified in saying"}}, "super_glue_copa": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name=None, template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.66}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.77}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.8}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.81}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "copa", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.84}, "template_name": "plausible_alternatives"}}, "super_glue_rte": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.7906137184115524}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.8267148014440433}, "template_name": "MNLI crowdsource"}, "does it follow that": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='does it follow that', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.7942238267148014}, "template_name": "does it follow that"}, "guaranteed true": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='guaranteed true', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.776173285198556}, "template_name": "guaranteed true"}, "should assume": {"arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='should assume', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "rte", "dataset_name": "super_glue", "evaluation": {"accuracy": 0.7617328519855595}, "template_name": "should assume"}}, "winogrande_winogrande_xl": {"Replace": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.5588003157063931}, "template_name": "Replace"}, "True or False": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.5280189423835833}, "template_name": "True or False"}, "does underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.5651144435674822}, "template_name": "does underscore refer to"}, "stand for": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.5082872928176796}, "template_name": "stand for"}, "underscore refer to": {"arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "winogrande_xl", "dataset_name": "winogrande", "evaluation": {"accuracy": 0.5651144435674822}, "template_name": "underscore refer to"}}, "xcopa_id": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.46}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.7}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.73}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.72}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.67}, "template_name": "plausible_alternatives"}}, "xcopa_sw": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.6}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.55}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.54}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.51}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.52}, "template_name": "plausible_alternatives"}}, "xcopa_ta": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.59}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.56}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.6}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.57}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.62}, "template_name": "plausible_alternatives"}}, "xcopa_vi": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.53}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.72}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.72}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.7}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.71}, "template_name": "plausible_alternatives"}}, "xcopa_zh": {"C1 or C2? premise, so/because\u2026": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.67}, "template_name": "C1 or C2? premise, so/because\u2026"}, "best_option": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.7}, "template_name": "best_option"}, "cause_effect": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.8}, "template_name": "cause_effect"}, "i_am_hesitating": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.77}, "template_name": "i_am_hesitating"}, "plausible_alternatives": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.79}, "template_name": "plausible_alternatives"}}, "xnli_ar": {"GPT-3 style": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5558232931726907}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.42128514056224897}, "template_name": "MNLI crowdsource"}, "can we infer": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5148594377510041}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40562248995983935}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4927710843373494}, "template_name": "justified in saying"}}, "xnli_en": {"GPT-3 style": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5891566265060241}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.42610441767068274}, "template_name": "MNLI crowdsource"}, "can we infer": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5662650602409639}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4614457831325301}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "en", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5437751004016064}, "template_name": "justified in saying"}}, "xnli_es": {"GPT-3 style": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5734939759036145}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40923694779116465}, "template_name": "MNLI crowdsource"}, "can we infer": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5148594377510041}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.43132530120481927}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4610441767068273}, "template_name": "justified in saying"}}, "xnli_fr": {"GPT-3 style": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5666666666666667}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.42208835341365464}, "template_name": "MNLI crowdsource"}, "can we infer": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5385542168674698}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.39076305220883534}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5100401606425703}, "template_name": "justified in saying"}}, "xnli_hi": {"GPT-3 style": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5345381526104418}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.41124497991967873}, "template_name": "MNLI crowdsource"}, "can we infer": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4751004016064257}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40923694779116465}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4469879518072289}, "template_name": "justified in saying"}}, "xnli_sw": {"GPT-3 style": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4827309236947791}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40562248995983935}, "template_name": "MNLI crowdsource"}, "can we infer": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.44497991967871486}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.42289156626506025}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xnli", "evaluation": {"accuracy": 0.41124497991967873}, "template_name": "justified in saying"}}, "xnli_ur": {"GPT-3 style": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4947791164658635}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.39759036144578314}, "template_name": "MNLI crowdsource"}, "can we infer": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4502008032128514}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.39036144578313253}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ur", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40843373493975904}, "template_name": "justified in saying"}}, "xnli_vi": {"GPT-3 style": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5449799196787148}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.40401606425702813}, "template_name": "MNLI crowdsource"}, "can we infer": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.44779116465863456}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4650602409638554}, "template_name": "justified in saying"}}, "xnli_zh": {"GPT-3 style": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5429718875502008}, "template_name": "GPT-3 style"}, "MNLI crowdsource": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.3891566265060241}, "template_name": "MNLI crowdsource"}, "can we infer": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.5032128514056224}, "template_name": "can we infer"}, "guaranteed/possible/impossible": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.38072289156626504}, "template_name": "guaranteed/possible/impossible"}, "justified in saying": {"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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xnli", "evaluation": {"accuracy": 0.4706827309236948}, "template_name": "justified in saying"}}}
\ No newline at end of file
diff --git a/evaluation_l1/story_cloze/2016/Answer_Given_options/results.json b/evaluation_l1/story_cloze/2016/Answer_Given_options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..bc12fbe77a80a053c5eed4af6d4f94b14a58fac8
--- /dev/null
+++ b/evaluation_l1/story_cloze/2016/Answer_Given_options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "story_cloze",
+  "dataset_config_name": "2016",
+  "template_name": "Answer Given options",
+  "evaluation": {
+    "accuracy": 0.8524853019775521
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Answer Given options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/story_cloze/2016/Choose_Story_Ending/results.json b/evaluation_l1/story_cloze/2016/Choose_Story_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2a0035cde5af1d05aa6218652e04dd7af635558e
--- /dev/null
+++ b/evaluation_l1/story_cloze/2016/Choose_Story_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "story_cloze",
+  "dataset_config_name": "2016",
+  "template_name": "Choose Story Ending",
+  "evaluation": {
+    "accuracy": 0.8957776590058792
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Choose Story Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/story_cloze/2016/Generate_Ending/results.json b/evaluation_l1/story_cloze/2016/Generate_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e806bc7ee1a263558fc203b0ac5385ec77a68abe
--- /dev/null
+++ b/evaluation_l1/story_cloze/2016/Generate_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "story_cloze",
+  "dataset_config_name": "2016",
+  "template_name": "Generate Ending",
+  "evaluation": {
+    "accuracy": 0.709246392303581
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Generate Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/story_cloze/2016/Novel_Correct_Ending/results.json b/evaluation_l1/story_cloze/2016/Novel_Correct_Ending/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad80f193b34315890ca3d493f219209e25720bf0
--- /dev/null
+++ b/evaluation_l1/story_cloze/2016/Novel_Correct_Ending/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "story_cloze",
+  "dataset_config_name": "2016",
+  "template_name": "Novel Correct Ending",
+  "evaluation": {
+    "accuracy": 0.8888295029396045
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Novel Correct Ending', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/story_cloze/2016/Story_Continuation_and_Options/results.json b/evaluation_l1/story_cloze/2016/Story_Continuation_and_Options/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6b10c247f5761114c564f5bf9b8ca6151706c5d7
--- /dev/null
+++ b/evaluation_l1/story_cloze/2016/Story_Continuation_and_Options/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "story_cloze",
+  "dataset_config_name": "2016",
+  "template_name": "Story Continuation and Options",
+  "evaluation": {
+    "accuracy": 0.8850881881346874
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='2016', dataset_name='story_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Story Continuation and Options', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/cb/GPT-3_style/results.json b/evaluation_l1/super_glue/cb/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e88caf3c7b226a207a9912791ce8586649b6a2e8
--- /dev/null
+++ b/evaluation_l1/super_glue/cb/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "cb",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.8392857142857143
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/cb/MNLI_crowdsource/results.json b/evaluation_l1/super_glue/cb/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..019653782e1cd06049e5b706b25af7d186617d43
--- /dev/null
+++ b/evaluation_l1/super_glue/cb/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "cb",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.35714285714285715
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/cb/can_we_infer/results.json b/evaluation_l1/super_glue/cb/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..5b4601cfe5ea3cb7ab2c7fb549deb030b057a8ea
--- /dev/null
+++ b/evaluation_l1/super_glue/cb/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "cb",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.7857142857142857
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/cb/guaranteed_possible_impossible/results.json b/evaluation_l1/super_glue/cb/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3c083f8ad9fc8581c84080c566637c1ef6a574b3
--- /dev/null
+++ b/evaluation_l1/super_glue/cb/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "cb",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.5535714285714286
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/cb/justified_in_saying/results.json b/evaluation_l1/super_glue/cb/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7260dd8971cd4a67d651fa81a27919f3f1e5fe07
--- /dev/null
+++ b/evaluation_l1/super_glue/cb/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "cb",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.7142857142857143
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='cb', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/copa/C1_or_C2?_premise/results.json b/evaluation_l1/super_glue/copa/C1_or_C2?_premise/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..4c3b5850c841659be2f1079fa7b5c632e04c5eaa
--- /dev/null
+++ b/evaluation_l1/super_glue/copa/C1_or_C2?_premise/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "copa",
+  "template_name": "C1 or C2? premise, so/because\u2026",
+  "evaluation": {
+    "accuracy": 0.66
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name=None, template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/copa/best_option/results.json b/evaluation_l1/super_glue/copa/best_option/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..cbba802897e2718d94eaca1ecd34843a2ddcd1d5
--- /dev/null
+++ b/evaluation_l1/super_glue/copa/best_option/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "copa",
+  "template_name": "best_option",
+  "evaluation": {
+    "accuracy": 0.77
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/copa/cause_effect/results.json b/evaluation_l1/super_glue/copa/cause_effect/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..9430d172ee9dcd88e334286ec0fc4121ec0114d9
--- /dev/null
+++ b/evaluation_l1/super_glue/copa/cause_effect/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "copa",
+  "template_name": "cause_effect",
+  "evaluation": {
+    "accuracy": 0.8
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/copa/i_am_hesitating/results.json b/evaluation_l1/super_glue/copa/i_am_hesitating/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6f89a66124cbcc5a49ef97ffe0e73dcc262f2b2e
--- /dev/null
+++ b/evaluation_l1/super_glue/copa/i_am_hesitating/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "copa",
+  "template_name": "i_am_hesitating",
+  "evaluation": {
+    "accuracy": 0.81
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/copa/plausible_alternatives/results.json b/evaluation_l1/super_glue/copa/plausible_alternatives/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a5aad2abd3c3fe329d7167231d1f0d7a00bab759
--- /dev/null
+++ b/evaluation_l1/super_glue/copa/plausible_alternatives/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "copa",
+  "template_name": "plausible_alternatives",
+  "evaluation": {
+    "accuracy": 0.84
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='copa', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/rte/GPT-3_style/results.json b/evaluation_l1/super_glue/rte/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6f0f21d99a6515d49923aa44a3555649a5a1c70
--- /dev/null
+++ b/evaluation_l1/super_glue/rte/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "rte",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.7906137184115524
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/rte/MNLI_crowdsource/results.json b/evaluation_l1/super_glue/rte/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..19415d9be79a38f6043de7d04791de5cc4bd6b1b
--- /dev/null
+++ b/evaluation_l1/super_glue/rte/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "rte",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.8267148014440433
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/rte/does_it_follow_that/results.json b/evaluation_l1/super_glue/rte/does_it_follow_that/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d372e9fb43eba25aac1c7aebf4a8174f2ec79495
--- /dev/null
+++ b/evaluation_l1/super_glue/rte/does_it_follow_that/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "rte",
+  "template_name": "does it follow that",
+  "evaluation": {
+    "accuracy": 0.7942238267148014
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='does it follow that', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/rte/guaranteed_true/results.json b/evaluation_l1/super_glue/rte/guaranteed_true/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..644170197819bd48259a1cf22f12340c27cb66e4
--- /dev/null
+++ b/evaluation_l1/super_glue/rte/guaranteed_true/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "rte",
+  "template_name": "guaranteed true",
+  "evaluation": {
+    "accuracy": 0.776173285198556
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='guaranteed true', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/super_glue/rte/should_assume/results.json b/evaluation_l1/super_glue/rte/should_assume/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..af31cd812073aff28baff4f147a77e4f11c96ae6
--- /dev/null
+++ b/evaluation_l1/super_glue/rte/should_assume/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "super_glue",
+  "dataset_config_name": "rte",
+  "template_name": "should assume",
+  "evaluation": {
+    "accuracy": 0.7617328519855595
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='rte', dataset_name='super_glue', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='should assume', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/winogrande/winogrande_xl/Replace/results.json b/evaluation_l1/winogrande/winogrande_xl/Replace/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d49b83bf5e085e24d6dc105eaba9f0f021af4049
--- /dev/null
+++ b/evaluation_l1/winogrande/winogrande_xl/Replace/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "winogrande",
+  "dataset_config_name": "winogrande_xl",
+  "template_name": "Replace",
+  "evaluation": {
+    "accuracy": 0.5588003157063931
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='Replace', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/winogrande/winogrande_xl/True_or_False/results.json b/evaluation_l1/winogrande/winogrande_xl/True_or_False/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..64986025f0e3d5fee2f07385485137e563059134
--- /dev/null
+++ b/evaluation_l1/winogrande/winogrande_xl/True_or_False/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "winogrande",
+  "dataset_config_name": "winogrande_xl",
+  "template_name": "True or False",
+  "evaluation": {
+    "accuracy": 0.5280189423835833
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='True or False', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/winogrande/winogrande_xl/does_underscore_refer_to/results.json b/evaluation_l1/winogrande/winogrande_xl/does_underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..df5c11afcdc5dba943fc3d5ac0d5a0eb0da5442b
--- /dev/null
+++ b/evaluation_l1/winogrande/winogrande_xl/does_underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "winogrande",
+  "dataset_config_name": "winogrande_xl",
+  "template_name": "does underscore refer to",
+  "evaluation": {
+    "accuracy": 0.5651144435674822
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='does underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/winogrande/winogrande_xl/stand_for/results.json b/evaluation_l1/winogrande/winogrande_xl/stand_for/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..647e2857c0218f770c23fff801cace541e8fdf83
--- /dev/null
+++ b/evaluation_l1/winogrande/winogrande_xl/stand_for/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "winogrande",
+  "dataset_config_name": "winogrande_xl",
+  "template_name": "stand for",
+  "evaluation": {
+    "accuracy": 0.5082872928176796
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='stand for', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/winogrande/winogrande_xl/underscore_refer_to/results.json b/evaluation_l1/winogrande/winogrande_xl/underscore_refer_to/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..826e581f8690184e6cc5502e619730ee5927a9f5
--- /dev/null
+++ b/evaluation_l1/winogrande/winogrande_xl/underscore_refer_to/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "winogrande",
+  "dataset_config_name": "winogrande_xl",
+  "template_name": "underscore refer to",
+  "evaluation": {
+    "accuracy": 0.5651144435674822
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='winogrande_xl', dataset_name='winogrande', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name=None, template_name='underscore refer to', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/id/C1_or_C2?_premise/results.json b/evaluation_l1/xcopa/id/C1_or_C2?_premise/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..293fdb26f92984e0268401b1f1a14b028be2162f
--- /dev/null
+++ b/evaluation_l1/xcopa/id/C1_or_C2?_premise/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "C1 or C2? premise, so/because\u2026",
+  "evaluation": {
+    "accuracy": 0.46
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/id/best_option/results.json b/evaluation_l1/xcopa/id/best_option/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..376f92d4d78eeb841b35b92e71a0a764063c2184
--- /dev/null
+++ b/evaluation_l1/xcopa/id/best_option/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "best_option",
+  "evaluation": {
+    "accuracy": 0.7
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/id/cause_effect/results.json b/evaluation_l1/xcopa/id/cause_effect/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..eb6e7d90f444d9b8c2b91f303b89e1fcd30e9022
--- /dev/null
+++ b/evaluation_l1/xcopa/id/cause_effect/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "cause_effect",
+  "evaluation": {
+    "accuracy": 0.73
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/id/i_am_hesitating/results.json b/evaluation_l1/xcopa/id/i_am_hesitating/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d8d8315697a9d8b396bdcf751d5a1125676ff60
--- /dev/null
+++ b/evaluation_l1/xcopa/id/i_am_hesitating/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "i_am_hesitating",
+  "evaluation": {
+    "accuracy": 0.72
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/id/plausible_alternatives/results.json b/evaluation_l1/xcopa/id/plausible_alternatives/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..32f9c69ce80a887a576c10e9ff705584cbc7c06b
--- /dev/null
+++ b/evaluation_l1/xcopa/id/plausible_alternatives/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "plausible_alternatives",
+  "evaluation": {
+    "accuracy": 0.67
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/sw/C1_or_C2?_premise/results.json b/evaluation_l1/xcopa/sw/C1_or_C2?_premise/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..96f7c8b279f70afb01baa8eef9e50ca71882b6cc
--- /dev/null
+++ b/evaluation_l1/xcopa/sw/C1_or_C2?_premise/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "C1 or C2? premise, so/because\u2026",
+  "evaluation": {
+    "accuracy": 0.6
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/sw/best_option/results.json b/evaluation_l1/xcopa/sw/best_option/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..5111b0ad2d757cbdce85a1168f2fbe34d3c5e38d
--- /dev/null
+++ b/evaluation_l1/xcopa/sw/best_option/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "best_option",
+  "evaluation": {
+    "accuracy": 0.55
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/sw/cause_effect/results.json b/evaluation_l1/xcopa/sw/cause_effect/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0ad7e70a308353cf9d37764620d23de296526cc2
--- /dev/null
+++ b/evaluation_l1/xcopa/sw/cause_effect/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "cause_effect",
+  "evaluation": {
+    "accuracy": 0.54
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/sw/i_am_hesitating/results.json b/evaluation_l1/xcopa/sw/i_am_hesitating/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3d49f6c7b070a1968668977bf88c48e43f1323f7
--- /dev/null
+++ b/evaluation_l1/xcopa/sw/i_am_hesitating/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "i_am_hesitating",
+  "evaluation": {
+    "accuracy": 0.51
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/sw/plausible_alternatives/results.json b/evaluation_l1/xcopa/sw/plausible_alternatives/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b536b7058230d3d08e87e8f6b91aeb84a91dbbf3
--- /dev/null
+++ b/evaluation_l1/xcopa/sw/plausible_alternatives/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "plausible_alternatives",
+  "evaluation": {
+    "accuracy": 0.52
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/ta/C1_or_C2?_premise/results.json b/evaluation_l1/xcopa/ta/C1_or_C2?_premise/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..37b0615e662498d1b2d51ccf76f87ed84baf4d4d
--- /dev/null
+++ b/evaluation_l1/xcopa/ta/C1_or_C2?_premise/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "C1 or C2? premise, so/because\u2026",
+  "evaluation": {
+    "accuracy": 0.59
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/ta/best_option/results.json b/evaluation_l1/xcopa/ta/best_option/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ad0d3fb1e2334d8f5a58ca45fdbc52234a155c5a
--- /dev/null
+++ b/evaluation_l1/xcopa/ta/best_option/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "best_option",
+  "evaluation": {
+    "accuracy": 0.56
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/ta/cause_effect/results.json b/evaluation_l1/xcopa/ta/cause_effect/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d0767df2f080cde3ed0b36bb73af7a735d30e1e1
--- /dev/null
+++ b/evaluation_l1/xcopa/ta/cause_effect/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "cause_effect",
+  "evaluation": {
+    "accuracy": 0.6
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/ta/i_am_hesitating/results.json b/evaluation_l1/xcopa/ta/i_am_hesitating/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..11ad208825e71b0e4b8c55821fb1c06bb0d2b8f8
--- /dev/null
+++ b/evaluation_l1/xcopa/ta/i_am_hesitating/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "i_am_hesitating",
+  "evaluation": {
+    "accuracy": 0.57
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/ta/plausible_alternatives/results.json b/evaluation_l1/xcopa/ta/plausible_alternatives/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..650cd2908644df0288ff02176a27ba6ccdd8430d
--- /dev/null
+++ b/evaluation_l1/xcopa/ta/plausible_alternatives/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "plausible_alternatives",
+  "evaluation": {
+    "accuracy": 0.62
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/vi/C1_or_C2?_premise/results.json b/evaluation_l1/xcopa/vi/C1_or_C2?_premise/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..4822d0163e81a6188e592916c789a7edf82fc895
--- /dev/null
+++ b/evaluation_l1/xcopa/vi/C1_or_C2?_premise/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "C1 or C2? premise, so/because\u2026",
+  "evaluation": {
+    "accuracy": 0.53
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/vi/best_option/results.json b/evaluation_l1/xcopa/vi/best_option/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..205c5e683fbb3fc4634f9fbfd28097c65d60620b
--- /dev/null
+++ b/evaluation_l1/xcopa/vi/best_option/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "best_option",
+  "evaluation": {
+    "accuracy": 0.72
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/vi/cause_effect/results.json b/evaluation_l1/xcopa/vi/cause_effect/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b8675fb5e6b88fd2daee764a82897ff98fc80d33
--- /dev/null
+++ b/evaluation_l1/xcopa/vi/cause_effect/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "cause_effect",
+  "evaluation": {
+    "accuracy": 0.72
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/vi/i_am_hesitating/results.json b/evaluation_l1/xcopa/vi/i_am_hesitating/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7a1e8a30a979e9b007efbf5ea69c4308305c6979
--- /dev/null
+++ b/evaluation_l1/xcopa/vi/i_am_hesitating/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "i_am_hesitating",
+  "evaluation": {
+    "accuracy": 0.7
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/vi/plausible_alternatives/results.json b/evaluation_l1/xcopa/vi/plausible_alternatives/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a754e5ef467a38023545605ed176de98e200240c
--- /dev/null
+++ b/evaluation_l1/xcopa/vi/plausible_alternatives/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "plausible_alternatives",
+  "evaluation": {
+    "accuracy": 0.71
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/zh/C1_or_C2?_premise/results.json b/evaluation_l1/xcopa/zh/C1_or_C2?_premise/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8785356a62493902dc90f963cf0c4e512544d145
--- /dev/null
+++ b/evaluation_l1/xcopa/zh/C1_or_C2?_premise/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "C1 or C2? premise, so/because\u2026",
+  "evaluation": {
+    "accuracy": 0.67
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='so/because\u2026,validation', target_max_length=256, template_config_name='en', template_name='C1 or C2? premise', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/zh/best_option/results.json b/evaluation_l1/xcopa/zh/best_option/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..66cc8b6be554597540a596a1e28b9e7b0322a4f6
--- /dev/null
+++ b/evaluation_l1/xcopa/zh/best_option/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "best_option",
+  "evaluation": {
+    "accuracy": 0.7
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='best_option', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/zh/cause_effect/results.json b/evaluation_l1/xcopa/zh/cause_effect/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6d6d9852b695c1d3c8725ca07fe44727f9072ce6
--- /dev/null
+++ b/evaluation_l1/xcopa/zh/cause_effect/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "cause_effect",
+  "evaluation": {
+    "accuracy": 0.8
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='cause_effect', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/zh/i_am_hesitating/results.json b/evaluation_l1/xcopa/zh/i_am_hesitating/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdd91c8fe51bf2cd88cecc831c3e10fb23956bb9
--- /dev/null
+++ b/evaluation_l1/xcopa/zh/i_am_hesitating/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "i_am_hesitating",
+  "evaluation": {
+    "accuracy": 0.77
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='i_am_hesitating', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xcopa/zh/plausible_alternatives/results.json b/evaluation_l1/xcopa/zh/plausible_alternatives/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2c703eb186efceb2023af78cdd7bb346f3c129c0
--- /dev/null
+++ b/evaluation_l1/xcopa/zh/plausible_alternatives/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "plausible_alternatives",
+  "evaluation": {
+    "accuracy": 0.79
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='plausible_alternatives', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ar/GPT-3_style/results.json b/evaluation_l1/xnli/ar/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..546639690a91af2f7657d5e3d11086607d70d92e
--- /dev/null
+++ b/evaluation_l1/xnli/ar/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.5558232931726907
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ar/MNLI_crowdsource/results.json b/evaluation_l1/xnli/ar/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..c5ca5a5de45223fdf04ffb89b42bd5210a6b05f4
--- /dev/null
+++ b/evaluation_l1/xnli/ar/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.42128514056224897
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ar/can_we_infer/results.json b/evaluation_l1/xnli/ar/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b4e668e7b7ac4839f822b30191a815ac8c574928
--- /dev/null
+++ b/evaluation_l1/xnli/ar/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.5148594377510041
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ar/guaranteed_possible_impossible/results.json b/evaluation_l1/xnli/ar/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e912b2ece40d36999fd229f24fc8ef3b329cc1a6
--- /dev/null
+++ b/evaluation_l1/xnli/ar/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.40562248995983935
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ar/justified_in_saying/results.json b/evaluation_l1/xnli/ar/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..fefe3744fd1e94bc92a8643385c9ca4150c0ddf6
--- /dev/null
+++ b/evaluation_l1/xnli/ar/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ar",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.4927710843373494
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/en/GPT-3_style/results.json b/evaluation_l1/xnli/en/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ce728dd89ed583d89ff66c427bc1e26e4c94b7c2
--- /dev/null
+++ b/evaluation_l1/xnli/en/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "en",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.5891566265060241
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/en/MNLI_crowdsource/results.json b/evaluation_l1/xnli/en/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b501c0c19d450e9ee0010114f311cb4265b37242
--- /dev/null
+++ b/evaluation_l1/xnli/en/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "en",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.42610441767068274
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/en/can_we_infer/results.json b/evaluation_l1/xnli/en/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..606887a2517b3583f762bbe473a0b0edebabbb4a
--- /dev/null
+++ b/evaluation_l1/xnli/en/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "en",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.5662650602409639
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/en/guaranteed_possible_impossible/results.json b/evaluation_l1/xnli/en/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8bb4647db40a2b53df96ab55502c1b458a16ba63
--- /dev/null
+++ b/evaluation_l1/xnli/en/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "en",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.4614457831325301
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/en/justified_in_saying/results.json b/evaluation_l1/xnli/en/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..fdf68b1b3c2db3ef8c9dd6d81fcb17a89e736e5a
--- /dev/null
+++ b/evaluation_l1/xnli/en/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "en",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.5437751004016064
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='en', dataset_name='xnli', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/es/GPT-3_style/results.json b/evaluation_l1/xnli/es/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..fbb0fbe72ce9a2366a5fbdfd9da6eb0f9d9fd040
--- /dev/null
+++ b/evaluation_l1/xnli/es/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.5734939759036145
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/es/MNLI_crowdsource/results.json b/evaluation_l1/xnli/es/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..9ad337da8b03428488d1855d4bb5512047f8b8b2
--- /dev/null
+++ b/evaluation_l1/xnli/es/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.40923694779116465
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/es/can_we_infer/results.json b/evaluation_l1/xnli/es/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c5b9dc83a39b27ee74ebc89a068c2f2257f6bb7
--- /dev/null
+++ b/evaluation_l1/xnli/es/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.5148594377510041
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/es/guaranteed_possible_impossible/results.json b/evaluation_l1/xnli/es/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6ad61caa2724825a2fe37d749ad9a5a0049b367c
--- /dev/null
+++ b/evaluation_l1/xnli/es/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.43132530120481927
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/es/justified_in_saying/results.json b/evaluation_l1/xnli/es/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..88a666927b00d9871cca5682f0ed58f889592173
--- /dev/null
+++ b/evaluation_l1/xnli/es/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "es",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.4610441767068273
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/fr/GPT-3_style/results.json b/evaluation_l1/xnli/fr/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b7406b5bc3ed6a427267f31abae653fa9308c394
--- /dev/null
+++ b/evaluation_l1/xnli/fr/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.5666666666666667
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/fr/MNLI_crowdsource/results.json b/evaluation_l1/xnli/fr/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..d4d16731dcbfbb4413466d1be513e13815368d38
--- /dev/null
+++ b/evaluation_l1/xnli/fr/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.42208835341365464
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/fr/can_we_infer/results.json b/evaluation_l1/xnli/fr/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..42b889b3a9d7148ac2c31a64cdc3eff07565c38a
--- /dev/null
+++ b/evaluation_l1/xnli/fr/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.5385542168674698
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/fr/guaranteed_possible_impossible/results.json b/evaluation_l1/xnli/fr/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1fb16a2540011994a7c5dec4cf07dad0c3ff536b
--- /dev/null
+++ b/evaluation_l1/xnli/fr/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.39076305220883534
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/fr/justified_in_saying/results.json b/evaluation_l1/xnli/fr/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..4f42e8570f20fcf1c468a2695865355610b7e971
--- /dev/null
+++ b/evaluation_l1/xnli/fr/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "fr",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.5100401606425703
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/hi/GPT-3_style/results.json b/evaluation_l1/xnli/hi/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3ff8a878683ceacf02f36c301dd43ce7bee51134
--- /dev/null
+++ b/evaluation_l1/xnli/hi/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.5345381526104418
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/hi/MNLI_crowdsource/results.json b/evaluation_l1/xnli/hi/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..46fd2c4c63efd9ffb52ca3cfcd385a92f730b432
--- /dev/null
+++ b/evaluation_l1/xnli/hi/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.41124497991967873
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/hi/can_we_infer/results.json b/evaluation_l1/xnli/hi/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ef0c2d34b4062ad7db36cc80415d5e6b588e4577
--- /dev/null
+++ b/evaluation_l1/xnli/hi/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.4751004016064257
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/hi/guaranteed_possible_impossible/results.json b/evaluation_l1/xnli/hi/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1637ac7ae1350312fe7e9291d2416cd3e50a9c79
--- /dev/null
+++ b/evaluation_l1/xnli/hi/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.40923694779116465
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/hi/justified_in_saying/results.json b/evaluation_l1/xnli/hi/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..951e5f90e621992983980077a06098e556e294af
--- /dev/null
+++ b/evaluation_l1/xnli/hi/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "hi",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.4469879518072289
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/sw/GPT-3_style/results.json b/evaluation_l1/xnli/sw/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..c819b498e7f52c050ecde967d689ef49c5ead831
--- /dev/null
+++ b/evaluation_l1/xnli/sw/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.4827309236947791
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/sw/MNLI_crowdsource/results.json b/evaluation_l1/xnli/sw/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..65398c49b780d2e20b7069e487311b6b981d4623
--- /dev/null
+++ b/evaluation_l1/xnli/sw/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.40562248995983935
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/sw/can_we_infer/results.json b/evaluation_l1/xnli/sw/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..39921e787dd0094bbc08607bf47d2937abbfd42a
--- /dev/null
+++ b/evaluation_l1/xnli/sw/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.44497991967871486
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/sw/guaranteed_possible_impossible/results.json b/evaluation_l1/xnli/sw/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..318bbebe5b8253865063239ce33e84307900a27c
--- /dev/null
+++ b/evaluation_l1/xnli/sw/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.42289156626506025
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/sw/justified_in_saying/results.json b/evaluation_l1/xnli/sw/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..cfff418d842a01be4a3654a50c02f6289a1fc703
--- /dev/null
+++ b/evaluation_l1/xnli/sw/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "sw",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.41124497991967873
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ur/GPT-3_style/results.json b/evaluation_l1/xnli/ur/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..24b3b4d9b3e429b610fb0a95c7722f51f1d0550a
--- /dev/null
+++ b/evaluation_l1/xnli/ur/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.4947791164658635
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ur/MNLI_crowdsource/results.json b/evaluation_l1/xnli/ur/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..67148069721a4c14b3c11b50b2aa8a4c09fe355b
--- /dev/null
+++ b/evaluation_l1/xnli/ur/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.39759036144578314
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ur/can_we_infer/results.json b/evaluation_l1/xnli/ur/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..22e46c6c09a04bf78053dd339ab28489f7704108
--- /dev/null
+++ b/evaluation_l1/xnli/ur/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.4502008032128514
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ur/guaranteed_possible_impossible/results.json b/evaluation_l1/xnli/ur/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b810a40e3e235790207376b32af2a299d4f6fe24
--- /dev/null
+++ b/evaluation_l1/xnli/ur/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.39036144578313253
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/ur/justified_in_saying/results.json b/evaluation_l1/xnli/ur/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..20245923546de8f7e85e8dd7e2bc1db0847212c3
--- /dev/null
+++ b/evaluation_l1/xnli/ur/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "ur",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.40843373493975904
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/vi/GPT-3_style/results.json b/evaluation_l1/xnli/vi/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a13ae51ee9460f9666c435cf5fa50778f44d72bb
--- /dev/null
+++ b/evaluation_l1/xnli/vi/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.5449799196787148
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/vi/MNLI_crowdsource/results.json b/evaluation_l1/xnli/vi/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ccbec33b9e9b2ed4137fee347c4b91c0d2da5e34
--- /dev/null
+++ b/evaluation_l1/xnli/vi/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.40401606425702813
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/vi/can_we_infer/results.json b/evaluation_l1/xnli/vi/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..92a1a38226419e2f25f841111418028bc08cc0cf
--- /dev/null
+++ b/evaluation_l1/xnli/vi/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.5
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/vi/guaranteed_possible_impossible/results.json b/evaluation_l1/xnli/vi/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..afde16840d9ce35ba02f3c2fcd7460a447b59b6a
--- /dev/null
+++ b/evaluation_l1/xnli/vi/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.44779116465863456
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/vi/justified_in_saying/results.json b/evaluation_l1/xnli/vi/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1afbfdbf7c938a67dd496eb79cde783df539b46a
--- /dev/null
+++ b/evaluation_l1/xnli/vi/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "vi",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.4650602409638554
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/zh/GPT-3_style/results.json b/evaluation_l1/xnli/zh/GPT-3_style/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2bb9b56add737900ef37773f0080e24295befdd1
--- /dev/null
+++ b/evaluation_l1/xnli/zh/GPT-3_style/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "GPT-3 style",
+  "evaluation": {
+    "accuracy": 0.5429718875502008
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='GPT-3 style', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/zh/MNLI_crowdsource/results.json b/evaluation_l1/xnli/zh/MNLI_crowdsource/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7dcfbc8f34fbd0614d389653502b05cc134b2cd7
--- /dev/null
+++ b/evaluation_l1/xnli/zh/MNLI_crowdsource/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "MNLI crowdsource",
+  "evaluation": {
+    "accuracy": 0.3891566265060241
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='MNLI crowdsource', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/zh/can_we_infer/results.json b/evaluation_l1/xnli/zh/can_we_infer/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ed0c5d7b69442e3a1dc012e67cc6dc92f926cb83
--- /dev/null
+++ b/evaluation_l1/xnli/zh/can_we_infer/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "can we infer",
+  "evaluation": {
+    "accuracy": 0.5032128514056224
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='can we infer', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/zh/guaranteed_possible_impossible/results.json b/evaluation_l1/xnli/zh/guaranteed_possible_impossible/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ab2e4c3507538fdc53545cdaa4b371ce69faba4f
--- /dev/null
+++ b/evaluation_l1/xnli/zh/guaranteed_possible_impossible/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "guaranteed/possible/impossible",
+  "evaluation": {
+    "accuracy": 0.38072289156626504
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='guaranteed/possible/impossible', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_l1/xnli/zh/justified_in_saying/results.json b/evaluation_l1/xnli/zh/justified_in_saying/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3e5f64a5ee3df4e2035281194d7f24398a2cb0fe
--- /dev/null
+++ b/evaluation_l1/xnli/zh/justified_in_saying/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xnli",
+  "dataset_config_name": "zh",
+  "template_name": "justified in saying",
+  "evaluation": {
+    "accuracy": 0.4706827309236948
+  },
+  "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/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='en', template_name='justified in saying', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Answer_Given_options_armt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Answer_Given_options_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..00b0919de5c779c5aded5f7884023a3e2b61f7ad
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Answer_Given_options_armt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Answer Given options_armt",
+  "evaluation": {
+    "accuracy": 0.6664460622104567
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Answer Given options_armt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Choose_Story_Ending_armt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Choose_Story_Ending_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a18875179d57b9179bf7e77c0044a588a4da04a6
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Choose_Story_Ending_armt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Choose Story Ending_armt",
+  "evaluation": {
+    "accuracy": 0.8385175380542687
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Choose Story Ending_armt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Generate_Ending_armt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Generate_Ending_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..417a9cacc699edaa2406c8aed141e11bf6f02342
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Generate_Ending_armt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Generate Ending_armt",
+  "evaluation": {
+    "accuracy": 0.5843812045003309
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Generate Ending_armt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Novel_Correct_Ending_armt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Novel_Correct_Ending_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3dddd8a167406b0f6a4f128561262f51a6da0a0d
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Novel_Correct_Ending_armt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Novel Correct Ending_armt",
+  "evaluation": {
+    "accuracy": 0.827928524156188
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Novel Correct Ending_armt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Story_Continuation_and_Options_armt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Story_Continuation_and_Options_armt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3dc021bd784256e5b7ad89410662d4bbe23814e3
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/ar/Story_Continuation_and_Options_armt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "ar",
+  "template_name": "Story Continuation and Options_armt",
+  "evaluation": {
+    "accuracy": 0.8246194573130378
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Story Continuation and Options_armt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Answer_Given_options_esmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Answer_Given_options_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2f80fdef48d02a647d1a396f1a1184b9de98f54d
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Answer_Given_options_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Answer Given options_esmt",
+  "evaluation": {
+    "accuracy": 0.8325612177365983
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Answer Given options_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Choose_Story_Ending_esmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Choose_Story_Ending_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ac33228a2c44794760c0541a9deca5bca51ddc3f
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Choose_Story_Ending_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Choose Story Ending_esmt",
+  "evaluation": {
+    "accuracy": 0.8881535407015222
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Choose Story Ending_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Generate_Ending_esmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Generate_Ending_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ef96b59e5aaf8917c3133bbbfdcda66635648262
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Generate_Ending_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Generate Ending_esmt",
+  "evaluation": {
+    "accuracy": 0.6776968894771674
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Generate Ending_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Novel_Correct_Ending_esmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Novel_Correct_Ending_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..592e9750b5c5381a60ad2a9b0e637323d4c28a2d
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Novel_Correct_Ending_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Novel Correct Ending_esmt",
+  "evaluation": {
+    "accuracy": 0.8656518861681006
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Novel Correct Ending_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Story_Continuation_and_Options_esmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Story_Continuation_and_Options_esmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..472c5c2061df541500d7e2e9e6eef586123bcea7
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/es/Story_Continuation_and_Options_esmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "es",
+  "template_name": "Story Continuation and Options_esmt",
+  "evaluation": {
+    "accuracy": 0.886168100595632
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Story Continuation and Options_esmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Answer_Given_options_eumt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Answer_Given_options_eumt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..c5f4b5d3ddd16a50e5746e8a401ebfef6af241af
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Answer_Given_options_eumt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Answer Given options_eumt",
+  "evaluation": {
+    "accuracy": 0.5678358702845797
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Answer Given options_eumt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Choose_Story_Ending_eumt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Choose_Story_Ending_eumt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8da88a477ea78793b373952fb1be96f481e158fe
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Choose_Story_Ending_eumt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Choose Story Ending_eumt",
+  "evaluation": {
+    "accuracy": 0.7326273990734613
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Choose Story Ending_eumt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Generate_Ending_eumt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Generate_Ending_eumt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..be084bbe7b32af5b5443df6df39bfdb1d06222b9
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Generate_Ending_eumt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Generate Ending_eumt",
+  "evaluation": {
+    "accuracy": 0.5095962938451357
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Generate Ending_eumt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Novel_Correct_Ending_eumt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Novel_Correct_Ending_eumt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ffd920c5fe19f4120e2f7b8cb9b03e75b4922a30
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Novel_Correct_Ending_eumt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Novel Correct Ending_eumt",
+  "evaluation": {
+    "accuracy": 0.6558570483123759
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Novel Correct Ending_eumt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Story_Continuation_and_Options_eumt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Story_Continuation_and_Options_eumt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..28733e61a32f89b95d611c458ff107eeaf7fa921
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/eu/Story_Continuation_and_Options_eumt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "eu",
+  "template_name": "Story Continuation and Options_eumt",
+  "evaluation": {
+    "accuracy": 0.7193911317008603
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Story Continuation and Options_eumt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Answer_Given_options_himt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Answer_Given_options_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c0aab623963b0e207dfa345e07433e8f4b2a462
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Answer_Given_options_himt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Answer Given options_himt",
+  "evaluation": {
+    "accuracy": 0.7054930509596293
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Answer Given options_himt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Choose_Story_Ending_himt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Choose_Story_Ending_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..9a75b90868fbec40451f3664a316ab82d5ba43e7
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Choose_Story_Ending_himt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Choose Story Ending_himt",
+  "evaluation": {
+    "accuracy": 0.8041032428855063
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Choose Story Ending_himt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Generate_Ending_himt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Generate_Ending_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..911d012a6b28cd79a12173b18896de370e018fd8
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Generate_Ending_himt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Generate Ending_himt",
+  "evaluation": {
+    "accuracy": 0.614824619457313
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Generate Ending_himt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Novel_Correct_Ending_himt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Novel_Correct_Ending_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..bdfdab2b619c12c446422d1160f1eaa296b71b61
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Novel_Correct_Ending_himt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Novel Correct Ending_himt",
+  "evaluation": {
+    "accuracy": 0.7584381204500331
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Novel Correct Ending_himt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Story_Continuation_and_Options_himt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Story_Continuation_and_Options_himt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1f4e6dd2de684bd0e4bcb456a7484fa808f7339c
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/hi/Story_Continuation_and_Options_himt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "hi",
+  "template_name": "Story Continuation and Options_himt",
+  "evaluation": {
+    "accuracy": 0.7981469225678358
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Story Continuation and Options_himt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Answer_Given_options_idmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Answer_Given_options_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6618121e00b79e2467033caee48503ab2b22c2cf
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Answer_Given_options_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Answer Given options_idmt",
+  "evaluation": {
+    "accuracy": 0.7326273990734613
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Answer Given options_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Choose_Story_Ending_idmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Choose_Story_Ending_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a3aa3e18e51c6b62cbb7cc2a97700041334c7e25
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Choose_Story_Ending_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Choose Story Ending_idmt",
+  "evaluation": {
+    "accuracy": 0.8457974851091992
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Choose Story Ending_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Generate_Ending_idmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Generate_Ending_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..c7b80d028191f714bf9413a37c734491b302092d
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Generate_Ending_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Generate Ending_idmt",
+  "evaluation": {
+    "accuracy": 0.5678358702845797
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Generate Ending_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Novel_Correct_Ending_idmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Novel_Correct_Ending_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3ea18beca2d084a68d50e0be3f03c41be93f2262
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Novel_Correct_Ending_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Novel Correct Ending_idmt",
+  "evaluation": {
+    "accuracy": 0.8226340172071476
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Novel Correct Ending_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Story_Continuation_and_Options_idmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Story_Continuation_and_Options_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e0984dea347a6d4a481eb42b9dc0c80f26693c89
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/id/Story_Continuation_and_Options_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "id",
+  "template_name": "Story Continuation and Options_idmt",
+  "evaluation": {
+    "accuracy": 0.8246194573130378
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Story Continuation and Options_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Answer_Given_options_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Answer_Given_options_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..e6ecf52678671b31c2c202352c3f9ee45d2103ff
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Answer_Given_options_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Answer Given options_zhmt",
+  "evaluation": {
+    "accuracy": 0.7935142289874255
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Answer Given options_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Choose_Story_Ending_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Choose_Story_Ending_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8254a6ef94f7708021f814ae35ccb144ccf766b2
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Choose_Story_Ending_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Choose Story Ending_zhmt",
+  "evaluation": {
+    "accuracy": 0.8590337524818001
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Choose Story Ending_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Generate_Ending_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Generate_Ending_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..86be5b81897365d846fdbc1c0212f1ca30b86bff
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Generate_Ending_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Generate Ending_zhmt",
+  "evaluation": {
+    "accuracy": 0.6307081403044341
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Generate Ending_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Novel_Correct_Ending_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Novel_Correct_Ending_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..a022479727e3f95ddd43e7725b85deff75157f3a
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Novel_Correct_Ending_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Novel Correct Ending_zhmt",
+  "evaluation": {
+    "accuracy": 0.8590337524818001
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Novel Correct Ending_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Story_Continuation_and_Options_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Story_Continuation_and_Options_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8a07d07dfc971f75f05a047a59fa74a581aae831
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xstory_cloze/zh/Story_Continuation_and_Options_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xstory_cloze",
+  "dataset_config_name": "zh",
+  "template_name": "Story Continuation and Options_zhmt",
+  "evaluation": {
+    "accuracy": 0.8464592984778293
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Story Continuation and Options_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/Replace_frmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/Replace_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7c1c9dbf4c2515bb812f95d4e251685bc13e5b98
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/Replace_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "Replace_frmt",
+  "evaluation": {
+    "accuracy": 0.5542168674698795
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='Replace_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/True_or_False_frmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/True_or_False_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..67bf5b6066b7e1974c71a7431bc80f60f6a36a99
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/True_or_False_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "True or False_frmt",
+  "evaluation": {
+    "accuracy": 0.46987951807228917
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='True or False_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/does_underscore_refer_to_frmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/does_underscore_refer_to_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..5be30d949e856fd08b7d57ec7e75260ecce3b1d4
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/does_underscore_refer_to_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "does underscore refer to_frmt",
+  "evaluation": {
+    "accuracy": 0.5301204819277109
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='does underscore refer to_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/stand_for_frmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/stand_for_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..aec05834b002a9c8c05edfa69aa43246be7035fd
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/stand_for_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "stand for_frmt",
+  "evaluation": {
+    "accuracy": 0.5662650602409639
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='stand for_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/underscore_refer_to_frmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/underscore_refer_to_frmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3be3c52fa7e754148451902755693f0495f4e710
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/fr/underscore_refer_to_frmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "fr",
+  "template_name": "underscore refer to_frmt",
+  "evaluation": {
+    "accuracy": 0.5783132530120482
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='underscore refer to_frmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/Replace_ptmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/Replace_ptmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..949c31b60675cdcef7fd55b20c28040f16d3882b
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/Replace_ptmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "Replace_ptmt",
+  "evaluation": {
+    "accuracy": 0.5551330798479087
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='Replace_ptmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/True_or_False_ptmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/True_or_False_ptmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..ca312b24a35e6eee10689ce9b36562e3525a6146
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/True_or_False_ptmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "True or False_ptmt",
+  "evaluation": {
+    "accuracy": 0.4600760456273764
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='True or False_ptmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/does_underscore_refer_to_ptmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/does_underscore_refer_to_ptmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..6bc1ed74c203359004d055de6ccfb7029966c693
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/does_underscore_refer_to_ptmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "does underscore refer to_ptmt",
+  "evaluation": {
+    "accuracy": 0.5513307984790875
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='does underscore refer to_ptmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/stand_for_ptmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/stand_for_ptmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..7aaa161d306ac5cf9ac2dbc18dfa356c209a237f
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/stand_for_ptmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "stand for_ptmt",
+  "evaluation": {
+    "accuracy": 0.532319391634981
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='stand for_ptmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/underscore_refer_to_ptmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/underscore_refer_to_ptmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..96182d824f2bb313ce572d25990c7d647e6d7177
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/pt/underscore_refer_to_ptmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "pt",
+  "template_name": "underscore refer to_ptmt",
+  "evaluation": {
+    "accuracy": 0.5361216730038023
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='underscore refer to_ptmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/Replace_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/Replace_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3f844872631efc4a4bdae4466528d694216e5834
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/Replace_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "Replace_zhmt",
+  "evaluation": {
+    "accuracy": 0.6130952380952381
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='Replace_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/True_or_False_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/True_or_False_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3936a48ccce8997a4f8b4667cc4ba8bfbc3c0b49
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/True_or_False_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "True or False_zhmt",
+  "evaluation": {
+    "accuracy": 0.5416666666666666
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='True or False_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/does_underscore_refer_to_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/does_underscore_refer_to_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..3aa2985b0967db2c47e54ddaa98b4f757e85a5b5
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/does_underscore_refer_to_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "does underscore refer to_zhmt",
+  "evaluation": {
+    "accuracy": 0.5793650793650794
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='does underscore refer to_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/stand_for_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/stand_for_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0037df9b42fa6ab40eb5106ae130594838d95548
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/stand_for_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "stand for_zhmt",
+  "evaluation": {
+    "accuracy": 0.5158730158730159
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='stand for_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/underscore_refer_to_zhmt/results.json b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/underscore_refer_to_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..703bbb0e5938c8a1219a5452fb1692557ed43b4d
--- /dev/null
+++ b/evaluation_xwinstorycopamt/Muennighoff_xwinograd/zh/underscore_refer_to_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "Muennighoff/xwinograd",
+  "dataset_config_name": "zh",
+  "template_name": "underscore refer to_zhmt",
+  "evaluation": {
+    "accuracy": 0.625
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='underscore refer to_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/merged.csv b/evaluation_xwinstorycopamt/merged.csv
new file mode 100644
index 0000000000000000000000000000000000000000..4a9402f0137311e94081fe18b349de0b71ea3071
--- /dev/null
+++ b/evaluation_xwinstorycopamt/merged.csv
@@ -0,0 +1,86 @@
+dataset,prompt,metric,value
+xcopa_id,C1 or C2? premise_idmt,accuracy,0.52
+xcopa_id,best_option_idmt,accuracy,0.63
+xcopa_id,cause_effect_idmt,accuracy,0.64
+xcopa_id,i_am_hesitating_idmt,accuracy,0.66
+xcopa_id,plausible_alternatives_idmt,accuracy,0.71
+xcopa_id,median,accuracy,0.64
+xcopa_sw,C1 or C2? premise_swmt,accuracy,0.6
+xcopa_sw,best_option_swmt,accuracy,0.59
+xcopa_sw,cause_effect_swmt,accuracy,0.57
+xcopa_sw,i_am_hesitating_swmt,accuracy,0.61
+xcopa_sw,plausible_alternatives_swmt,accuracy,0.59
+xcopa_sw,median,accuracy,0.59
+xcopa_ta,C1 or C2? premise_tamt,accuracy,0.6
+xcopa_ta,best_option_tamt,accuracy,0.56
+xcopa_ta,cause_effect_tamt,accuracy,0.58
+xcopa_ta,i_am_hesitating_tamt,accuracy,0.54
+xcopa_ta,plausible_alternatives_tamt,accuracy,0.54
+xcopa_ta,median,accuracy,0.56
+xcopa_vi,C1 or C2? premise_vimt,accuracy,0.63
+xcopa_vi,best_option_vimt,accuracy,0.73
+xcopa_vi,cause_effect_vimt,accuracy,0.72
+xcopa_vi,i_am_hesitating_vimt,accuracy,0.71
+xcopa_vi,plausible_alternatives_vimt,accuracy,0.77
+xcopa_vi,median,accuracy,0.72
+xcopa_zh,C1 or C2? premise_zhmt,accuracy,0.61
+xcopa_zh,best_option_zhmt,accuracy,0.69
+xcopa_zh,cause_effect_zhmt,accuracy,0.8
+xcopa_zh,i_am_hesitating_zhmt,accuracy,0.74
+xcopa_zh,plausible_alternatives_zhmt,accuracy,0.76
+xcopa_zh,median,accuracy,0.74
+xstory_cloze_ar,Answer Given options_armt,accuracy,0.6664460622104567
+xstory_cloze_ar,Choose Story Ending_armt,accuracy,0.8385175380542687
+xstory_cloze_ar,Generate Ending_armt,accuracy,0.5843812045003309
+xstory_cloze_ar,Novel Correct Ending_armt,accuracy,0.827928524156188
+xstory_cloze_ar,Story Continuation and Options_armt,accuracy,0.8246194573130378
+xstory_cloze_ar,median,accuracy,0.8246194573130378
+xstory_cloze_es,Answer Given options_esmt,accuracy,0.8325612177365983
+xstory_cloze_es,Choose Story Ending_esmt,accuracy,0.8881535407015222
+xstory_cloze_es,Generate Ending_esmt,accuracy,0.6776968894771674
+xstory_cloze_es,Novel Correct Ending_esmt,accuracy,0.8656518861681006
+xstory_cloze_es,Story Continuation and Options_esmt,accuracy,0.886168100595632
+xstory_cloze_es,median,accuracy,0.8656518861681006
+xstory_cloze_eu,Answer Given options_eumt,accuracy,0.5678358702845797
+xstory_cloze_eu,Choose Story Ending_eumt,accuracy,0.7326273990734613
+xstory_cloze_eu,Generate Ending_eumt,accuracy,0.5095962938451357
+xstory_cloze_eu,Novel Correct Ending_eumt,accuracy,0.6558570483123759
+xstory_cloze_eu,Story Continuation and Options_eumt,accuracy,0.7193911317008603
+xstory_cloze_eu,median,accuracy,0.6558570483123759
+xstory_cloze_hi,Answer Given options_himt,accuracy,0.7054930509596293
+xstory_cloze_hi,Choose Story Ending_himt,accuracy,0.8041032428855063
+xstory_cloze_hi,Generate Ending_himt,accuracy,0.614824619457313
+xstory_cloze_hi,Novel Correct Ending_himt,accuracy,0.7584381204500331
+xstory_cloze_hi,Story Continuation and Options_himt,accuracy,0.7981469225678358
+xstory_cloze_hi,median,accuracy,0.7584381204500331
+xstory_cloze_id,Answer Given options_idmt,accuracy,0.7326273990734613
+xstory_cloze_id,Choose Story Ending_idmt,accuracy,0.8457974851091992
+xstory_cloze_id,Generate Ending_idmt,accuracy,0.5678358702845797
+xstory_cloze_id,Novel Correct Ending_idmt,accuracy,0.8226340172071476
+xstory_cloze_id,Story Continuation and Options_idmt,accuracy,0.8246194573130378
+xstory_cloze_id,median,accuracy,0.8226340172071476
+xstory_cloze_zh,Answer Given options_zhmt,accuracy,0.7935142289874255
+xstory_cloze_zh,Choose Story Ending_zhmt,accuracy,0.8590337524818001
+xstory_cloze_zh,Generate Ending_zhmt,accuracy,0.6307081403044341
+xstory_cloze_zh,Novel Correct Ending_zhmt,accuracy,0.8590337524818001
+xstory_cloze_zh,Story Continuation and Options_zhmt,accuracy,0.8464592984778293
+xstory_cloze_zh,median,accuracy,0.8464592984778293
+xwinograd_fr,Replace_frmt,accuracy,0.5542168674698795
+xwinograd_fr,True or False_frmt,accuracy,0.46987951807228917
+xwinograd_fr,does underscore refer to_frmt,accuracy,0.5301204819277109
+xwinograd_fr,stand for_frmt,accuracy,0.5662650602409639
+xwinograd_fr,underscore refer to_frmt,accuracy,0.5783132530120482
+xwinograd_fr,median,accuracy,0.5542168674698795
+xwinograd_pt,Replace_ptmt,accuracy,0.5551330798479087
+xwinograd_pt,True or False_ptmt,accuracy,0.4600760456273764
+xwinograd_pt,does underscore refer to_ptmt,accuracy,0.5513307984790875
+xwinograd_pt,stand for_ptmt,accuracy,0.532319391634981
+xwinograd_pt,underscore refer to_ptmt,accuracy,0.5361216730038023
+xwinograd_pt,median,accuracy,0.5361216730038023
+xwinograd_zh,Replace_zhmt,accuracy,0.6130952380952381
+xwinograd_zh,True or False_zhmt,accuracy,0.5416666666666666
+xwinograd_zh,does underscore refer to_zhmt,accuracy,0.5793650793650794
+xwinograd_zh,stand for_zhmt,accuracy,0.5158730158730159
+xwinograd_zh,underscore refer to_zhmt,accuracy,0.625
+xwinograd_zh,median,accuracy,0.5793650793650794
+multiple,average,multiple,0.692383103411949
diff --git a/evaluation_xwinstorycopamt/merged.json b/evaluation_xwinstorycopamt/merged.json
new file mode 100644
index 0000000000000000000000000000000000000000..623985f31ca9ea9cd805f20932eab97e9204faa9
--- /dev/null
+++ b/evaluation_xwinstorycopamt/merged.json
@@ -0,0 +1 @@
+{"Muennighoff/xstory_cloze_ar": {"Answer Given options_armt": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Answer Given options_armt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6664460622104567}, "template_name": "Answer Given options_armt"}, "Choose Story Ending_armt": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Choose Story Ending_armt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8385175380542687}, "template_name": "Choose Story Ending_armt"}, "Generate Ending_armt": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Generate Ending_armt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.5843812045003309}, "template_name": "Generate Ending_armt"}, "Novel Correct Ending_armt": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Novel Correct Ending_armt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.827928524156188}, "template_name": "Novel Correct Ending_armt"}, "Story Continuation and Options_armt": {"arguments": "Namespace(config_name=None, dataset_config_name='ar', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Story Continuation and Options_armt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ar", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8246194573130378}, "template_name": "Story Continuation and Options_armt"}}, "Muennighoff/xstory_cloze_es": {"Answer Given options_esmt": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Answer Given options_esmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8325612177365983}, "template_name": "Answer Given options_esmt"}, "Choose Story Ending_esmt": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Choose Story Ending_esmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8881535407015222}, "template_name": "Choose Story Ending_esmt"}, "Generate Ending_esmt": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Generate Ending_esmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6776968894771674}, "template_name": "Generate Ending_esmt"}, "Novel Correct Ending_esmt": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Novel Correct Ending_esmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8656518861681006}, "template_name": "Novel Correct Ending_esmt"}, "Story Continuation and Options_esmt": {"arguments": "Namespace(config_name=None, dataset_config_name='es', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Story Continuation and Options_esmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "es", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.886168100595632}, "template_name": "Story Continuation and Options_esmt"}}, "Muennighoff/xstory_cloze_eu": {"Answer Given options_eumt": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Answer Given options_eumt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.5678358702845797}, "template_name": "Answer Given options_eumt"}, "Choose Story Ending_eumt": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Choose Story Ending_eumt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7326273990734613}, "template_name": "Choose Story Ending_eumt"}, "Generate Ending_eumt": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Generate Ending_eumt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.5095962938451357}, "template_name": "Generate Ending_eumt"}, "Novel Correct Ending_eumt": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Novel Correct Ending_eumt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6558570483123759}, "template_name": "Novel Correct Ending_eumt"}, "Story Continuation and Options_eumt": {"arguments": "Namespace(config_name=None, dataset_config_name='eu', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='eu', template_name='Story Continuation and Options_eumt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "eu", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7193911317008603}, "template_name": "Story Continuation and Options_eumt"}}, "Muennighoff/xstory_cloze_hi": {"Answer Given options_himt": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Answer Given options_himt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7054930509596293}, "template_name": "Answer Given options_himt"}, "Choose Story Ending_himt": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Choose Story Ending_himt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8041032428855063}, "template_name": "Choose Story Ending_himt"}, "Generate Ending_himt": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Generate Ending_himt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.614824619457313}, "template_name": "Generate Ending_himt"}, "Novel Correct Ending_himt": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Novel Correct Ending_himt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7584381204500331}, "template_name": "Novel Correct Ending_himt"}, "Story Continuation and Options_himt": {"arguments": "Namespace(config_name=None, dataset_config_name='hi', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Story Continuation and Options_himt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "hi", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7981469225678358}, "template_name": "Story Continuation and Options_himt"}}, "Muennighoff/xstory_cloze_id": {"Answer Given options_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Answer Given options_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7326273990734613}, "template_name": "Answer Given options_idmt"}, "Choose Story Ending_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Choose Story Ending_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8457974851091992}, "template_name": "Choose Story Ending_idmt"}, "Generate Ending_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Generate Ending_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.5678358702845797}, "template_name": "Generate Ending_idmt"}, "Novel Correct Ending_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Novel Correct Ending_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8226340172071476}, "template_name": "Novel Correct Ending_idmt"}, "Story Continuation and Options_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='Story Continuation and Options_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8246194573130378}, "template_name": "Story Continuation and Options_idmt"}}, "Muennighoff/xstory_cloze_zh": {"Answer Given options_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Answer Given options_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.7935142289874255}, "template_name": "Answer Given options_zhmt"}, "Choose Story Ending_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Choose Story Ending_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8590337524818001}, "template_name": "Choose Story Ending_zhmt"}, "Generate Ending_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Generate Ending_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.6307081403044341}, "template_name": "Generate Ending_zhmt"}, "Novel Correct Ending_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Novel Correct Ending_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8590337524818001}, "template_name": "Novel Correct Ending_zhmt"}, "Story Continuation and Options_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xstory_cloze', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='Story Continuation and Options_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xstory_cloze", "evaluation": {"accuracy": 0.8464592984778293}, "template_name": "Story Continuation and Options_zhmt"}}, "Muennighoff/xwinograd_fr": {"Replace_frmt": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='Replace_frmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5542168674698795}, "template_name": "Replace_frmt"}, "True or False_frmt": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='True or False_frmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.46987951807228917}, "template_name": "True or False_frmt"}, "does underscore refer to_frmt": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='does underscore refer to_frmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5301204819277109}, "template_name": "does underscore refer to_frmt"}, "stand for_frmt": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='stand for_frmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5662650602409639}, "template_name": "stand for_frmt"}, "underscore refer to_frmt": {"arguments": "Namespace(config_name=None, dataset_config_name='fr', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='fr', template_name='underscore refer to_frmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "fr", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5783132530120482}, "template_name": "underscore refer to_frmt"}}, "Muennighoff/xwinograd_pt": {"Replace_ptmt": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='Replace_ptmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5551330798479087}, "template_name": "Replace_ptmt"}, "True or False_ptmt": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='True or False_ptmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.4600760456273764}, "template_name": "True or False_ptmt"}, "does underscore refer to_ptmt": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='does underscore refer to_ptmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5513307984790875}, "template_name": "does underscore refer to_ptmt"}, "stand for_ptmt": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='stand for_ptmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.532319391634981}, "template_name": "stand for_ptmt"}, "underscore refer to_ptmt": {"arguments": "Namespace(config_name=None, dataset_config_name='pt', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='pt', template_name='underscore refer to_ptmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "pt", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5361216730038023}, "template_name": "underscore refer to_ptmt"}}, "Muennighoff/xwinograd_zh": {"Replace_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='Replace_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.6130952380952381}, "template_name": "Replace_zhmt"}, "True or False_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='True or False_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5416666666666666}, "template_name": "True or False_zhmt"}, "does underscore refer to_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='does underscore refer to_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5793650793650794}, "template_name": "does underscore refer to_zhmt"}, "stand for_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='stand for_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.5158730158730159}, "template_name": "stand for_zhmt"}, "underscore refer to_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='Muennighoff/xwinograd', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='test', target_max_length=256, template_config_name='zh', template_name='underscore refer to_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "Muennighoff/xwinograd", "evaluation": {"accuracy": 0.625}, "template_name": "underscore refer to_zhmt"}}, "xcopa_id": {"C1 or C2? premise_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='C1 or C2? premise_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.52}, "template_name": "C1 or C2? premise_idmt"}, "best_option_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='best_option_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.63}, "template_name": "best_option_idmt"}, "cause_effect_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='cause_effect_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.64}, "template_name": "cause_effect_idmt"}, "i_am_hesitating_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='i_am_hesitating_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.66}, "template_name": "i_am_hesitating_idmt"}, "plausible_alternatives_idmt": {"arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='plausible_alternatives_idmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "id", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.71}, "template_name": "plausible_alternatives_idmt"}}, "xcopa_sw": {"C1 or C2? premise_swmt": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='C1 or C2? premise_swmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.6}, "template_name": "C1 or C2? premise_swmt"}, "best_option_swmt": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='best_option_swmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.59}, "template_name": "best_option_swmt"}, "cause_effect_swmt": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='cause_effect_swmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.57}, "template_name": "cause_effect_swmt"}, "i_am_hesitating_swmt": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='i_am_hesitating_swmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.61}, "template_name": "i_am_hesitating_swmt"}, "plausible_alternatives_swmt": {"arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='plausible_alternatives_swmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "sw", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.59}, "template_name": "plausible_alternatives_swmt"}}, "xcopa_ta": {"C1 or C2? premise_tamt": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='C1 or C2? premise_tamt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.6}, "template_name": "C1 or C2? premise_tamt"}, "best_option_tamt": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='best_option_tamt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.56}, "template_name": "best_option_tamt"}, "cause_effect_tamt": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='cause_effect_tamt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.58}, "template_name": "cause_effect_tamt"}, "i_am_hesitating_tamt": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='i_am_hesitating_tamt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.54}, "template_name": "i_am_hesitating_tamt"}, "plausible_alternatives_tamt": {"arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='plausible_alternatives_tamt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "ta", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.54}, "template_name": "plausible_alternatives_tamt"}}, "xcopa_vi": {"C1 or C2? premise_vimt": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='C1 or C2? premise_vimt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.63}, "template_name": "C1 or C2? premise_vimt"}, "best_option_vimt": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='best_option_vimt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.73}, "template_name": "best_option_vimt"}, "cause_effect_vimt": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='cause_effect_vimt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.72}, "template_name": "cause_effect_vimt"}, "i_am_hesitating_vimt": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='i_am_hesitating_vimt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.71}, "template_name": "i_am_hesitating_vimt"}, "plausible_alternatives_vimt": {"arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='plausible_alternatives_vimt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "vi", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.77}, "template_name": "plausible_alternatives_vimt"}}, "xcopa_zh": {"C1 or C2? premise_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='C1 or C2? premise_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.61}, "template_name": "C1 or C2? premise_zhmt"}, "best_option_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='best_option_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.69}, "template_name": "best_option_zhmt"}, "cause_effect_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='cause_effect_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.8}, "template_name": "cause_effect_zhmt"}, "i_am_hesitating_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='i_am_hesitating_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.74}, "template_name": "i_am_hesitating_zhmt"}, "plausible_alternatives_zhmt": {"arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='plausible_alternatives_zhmt', tokenizer_name=None, use_slow_tokenizer=False)", "dataset_config_name": "zh", "dataset_name": "xcopa", "evaluation": {"accuracy": 0.76}, "template_name": "plausible_alternatives_zhmt"}}}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/id/C1_or_C2?_premise_idmt/results.json b/evaluation_xwinstorycopamt/xcopa/id/C1_or_C2?_premise_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..770aecf4ada66a83385145616e7007789a7cb8d9
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/id/C1_or_C2?_premise_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "C1 or C2? premise_idmt",
+  "evaluation": {
+    "accuracy": 0.52
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='C1 or C2? premise_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/id/best_option_idmt/results.json b/evaluation_xwinstorycopamt/xcopa/id/best_option_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..423b21a96fd8c6769ab5ebcaac6e25e336722eaf
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/id/best_option_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "best_option_idmt",
+  "evaluation": {
+    "accuracy": 0.63
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='best_option_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/id/cause_effect_idmt/results.json b/evaluation_xwinstorycopamt/xcopa/id/cause_effect_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b4e8f8d7a3504693c80597ba0e247f27744daf0c
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/id/cause_effect_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "cause_effect_idmt",
+  "evaluation": {
+    "accuracy": 0.64
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='cause_effect_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/id/i_am_hesitating_idmt/results.json b/evaluation_xwinstorycopamt/xcopa/id/i_am_hesitating_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..aff6b8404ecd49a65dcf440e4a77ed71facf41cd
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/id/i_am_hesitating_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "i_am_hesitating_idmt",
+  "evaluation": {
+    "accuracy": 0.66
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='i_am_hesitating_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/id/plausible_alternatives_idmt/results.json b/evaluation_xwinstorycopamt/xcopa/id/plausible_alternatives_idmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..093ee47bb082a31a917c4f4e1bdffbfcff027208
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/id/plausible_alternatives_idmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "id",
+  "template_name": "plausible_alternatives_idmt",
+  "evaluation": {
+    "accuracy": 0.71
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='id', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='id', template_name='plausible_alternatives_idmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/sw/C1_or_C2?_premise_swmt/results.json b/evaluation_xwinstorycopamt/xcopa/sw/C1_or_C2?_premise_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0d0b7e0c2986140a20ac0007be6a9a7d9c7f3e06
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/sw/C1_or_C2?_premise_swmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "C1 or C2? premise_swmt",
+  "evaluation": {
+    "accuracy": 0.6
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='C1 or C2? premise_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/sw/best_option_swmt/results.json b/evaluation_xwinstorycopamt/xcopa/sw/best_option_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..30ec8563658088b618c110d1449b85a22476eb41
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/sw/best_option_swmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "best_option_swmt",
+  "evaluation": {
+    "accuracy": 0.59
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='best_option_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/sw/cause_effect_swmt/results.json b/evaluation_xwinstorycopamt/xcopa/sw/cause_effect_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..8695c2a6f0d6784b1b6df9924a65ba837756f004
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/sw/cause_effect_swmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "cause_effect_swmt",
+  "evaluation": {
+    "accuracy": 0.57
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='cause_effect_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/sw/i_am_hesitating_swmt/results.json b/evaluation_xwinstorycopamt/xcopa/sw/i_am_hesitating_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..1c30790a87601adf49bb4b88c549419044c85c1a
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/sw/i_am_hesitating_swmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "i_am_hesitating_swmt",
+  "evaluation": {
+    "accuracy": 0.61
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='i_am_hesitating_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/sw/plausible_alternatives_swmt/results.json b/evaluation_xwinstorycopamt/xcopa/sw/plausible_alternatives_swmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..cbb5a24fee689670d04bddd69bcae9af62589b0d
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/sw/plausible_alternatives_swmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "sw",
+  "template_name": "plausible_alternatives_swmt",
+  "evaluation": {
+    "accuracy": 0.59
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='sw', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='plausible_alternatives_swmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/ta/C1_or_C2?_premise_tamt/results.json b/evaluation_xwinstorycopamt/xcopa/ta/C1_or_C2?_premise_tamt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..28dc514e513e8365cf469b4c6bcb29b7a578e100
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/ta/C1_or_C2?_premise_tamt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "C1 or C2? premise_tamt",
+  "evaluation": {
+    "accuracy": 0.6
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='C1 or C2? premise_tamt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/ta/best_option_tamt/results.json b/evaluation_xwinstorycopamt/xcopa/ta/best_option_tamt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..78542a8227f63dcf1c8e7a433b7bb8027d262573
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/ta/best_option_tamt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "best_option_tamt",
+  "evaluation": {
+    "accuracy": 0.56
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='best_option_tamt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/ta/cause_effect_tamt/results.json b/evaluation_xwinstorycopamt/xcopa/ta/cause_effect_tamt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..0b63cfc7d5d6f767e20bed692b47acbf6644acda
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/ta/cause_effect_tamt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "cause_effect_tamt",
+  "evaluation": {
+    "accuracy": 0.58
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='cause_effect_tamt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/ta/i_am_hesitating_tamt/results.json b/evaluation_xwinstorycopamt/xcopa/ta/i_am_hesitating_tamt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..addeff34489644602e10a13a32136a853d11a027
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/ta/i_am_hesitating_tamt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "i_am_hesitating_tamt",
+  "evaluation": {
+    "accuracy": 0.54
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='i_am_hesitating_tamt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/ta/plausible_alternatives_tamt/results.json b/evaluation_xwinstorycopamt/xcopa/ta/plausible_alternatives_tamt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..956fbb245f1ce4cb95cabc8bb699a728716e7a49
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/ta/plausible_alternatives_tamt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "ta",
+  "template_name": "plausible_alternatives_tamt",
+  "evaluation": {
+    "accuracy": 0.54
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='ta', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/evaluation', pad_to_max_length=False, per_device_eval_batch_size=8, prefixlm=False, split='validation', target_max_length=256, template_config_name='ta', template_name='plausible_alternatives_tamt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/vi/C1_or_C2?_premise_vimt/results.json b/evaluation_xwinstorycopamt/xcopa/vi/C1_or_C2?_premise_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2028137a9042824411e47cae1ea23f520716e0c0
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/vi/C1_or_C2?_premise_vimt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "C1 or C2? premise_vimt",
+  "evaluation": {
+    "accuracy": 0.63
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='C1 or C2? premise_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/vi/best_option_vimt/results.json b/evaluation_xwinstorycopamt/xcopa/vi/best_option_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..67fb890f45a0d2e5054c512361c5454687ba31e1
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/vi/best_option_vimt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "best_option_vimt",
+  "evaluation": {
+    "accuracy": 0.73
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='best_option_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/vi/cause_effect_vimt/results.json b/evaluation_xwinstorycopamt/xcopa/vi/cause_effect_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..33fcb6b0693c93cd8908bca9e278290442c11cf0
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/vi/cause_effect_vimt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "cause_effect_vimt",
+  "evaluation": {
+    "accuracy": 0.72
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='cause_effect_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/vi/i_am_hesitating_vimt/results.json b/evaluation_xwinstorycopamt/xcopa/vi/i_am_hesitating_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..93cc0bc854520f7c93ebeb07b7d3d73ecc93cf19
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/vi/i_am_hesitating_vimt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "i_am_hesitating_vimt",
+  "evaluation": {
+    "accuracy": 0.71
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='i_am_hesitating_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/vi/plausible_alternatives_vimt/results.json b/evaluation_xwinstorycopamt/xcopa/vi/plausible_alternatives_vimt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..bdd4d1a4aa56e09ec45ee9b8f071edae2dd7450b
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/vi/plausible_alternatives_vimt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "vi",
+  "template_name": "plausible_alternatives_vimt",
+  "evaluation": {
+    "accuracy": 0.77
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='vi', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='plausible_alternatives_vimt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/zh/C1_or_C2?_premise_zhmt/results.json b/evaluation_xwinstorycopamt/xcopa/zh/C1_or_C2?_premise_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..2970afcaeb8d36e3a35214b3c4c778ac18752cf0
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/zh/C1_or_C2?_premise_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "C1 or C2? premise_zhmt",
+  "evaluation": {
+    "accuracy": 0.61
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='C1 or C2? premise_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/zh/best_option_zhmt/results.json b/evaluation_xwinstorycopamt/xcopa/zh/best_option_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..4cd6431f41ee122238b455e34f0bf4c95e383c5e
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/zh/best_option_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "best_option_zhmt",
+  "evaluation": {
+    "accuracy": 0.69
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='best_option_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/zh/cause_effect_zhmt/results.json b/evaluation_xwinstorycopamt/xcopa/zh/cause_effect_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..aa0aa530ab5f84188622f842b58fd663198c7bb7
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/zh/cause_effect_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "cause_effect_zhmt",
+  "evaluation": {
+    "accuracy": 0.8
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='cause_effect_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/zh/i_am_hesitating_zhmt/results.json b/evaluation_xwinstorycopamt/xcopa/zh/i_am_hesitating_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..b61d725b211203c13659dc312de7ca007207fa17
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/zh/i_am_hesitating_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "i_am_hesitating_zhmt",
+  "evaluation": {
+    "accuracy": 0.74
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='i_am_hesitating_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/evaluation_xwinstorycopamt/xcopa/zh/plausible_alternatives_zhmt/results.json b/evaluation_xwinstorycopamt/xcopa/zh/plausible_alternatives_zhmt/results.json
new file mode 100644
index 0000000000000000000000000000000000000000..049f76fe863b55ba2f0881cb3b146bba011f82cc
--- /dev/null
+++ b/evaluation_xwinstorycopamt/xcopa/zh/plausible_alternatives_zhmt/results.json
@@ -0,0 +1,9 @@
+{
+  "dataset_name": "xcopa",
+  "dataset_config_name": "zh",
+  "template_name": "plausible_alternatives_zhmt",
+  "evaluation": {
+    "accuracy": 0.76
+  },
+  "arguments": "Namespace(config_name=None, dataset_config_name='zh', dataset_name='xcopa', debug=False, dtype='float16', max_length=2048, model_name_or_path='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt', output_dir='/gpfsscratch/rech/six/commun/experiments/muennighoff/bloomckpt/6b3t0/tr13f-6b3-ml-t0-lmtoks341b-t0toks4b-xp3mt/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='plausible_alternatives_zhmt', tokenizer_name=None, use_slow_tokenizer=False)"
+}
\ No newline at end of file
diff --git a/pytorch_model.bin b/pytorch_model.bin
new file mode 100644
index 0000000000000000000000000000000000000000..d0a07bf673b31e491bbd79d9cb83a4f2d7a3411c
--- /dev/null
+++ b/pytorch_model.bin
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:915c66db82e1e36012dee6b134aeb9427ff3f722ae2a296806aea4f439a0f42a
+size 14138162687
diff --git a/special_tokens_map.json b/special_tokens_map.json
new file mode 100644
index 0000000000000000000000000000000000000000..25bc39604f72700b3b8e10bd69bb2f227157edd1
--- /dev/null
+++ b/special_tokens_map.json
@@ -0,0 +1 @@
+{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "pad_token": "<pad>"}
\ No newline at end of file
diff --git a/tokenizer.json b/tokenizer.json
new file mode 100644
index 0000000000000000000000000000000000000000..370bd68e20b4b6574ee05b213a74b244e3f492f3
--- /dev/null
+++ b/tokenizer.json
@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fa39cd4b1500feb205bcce3b9703a4373414cafe4970e0657b413f7ddd2a9d3
+size 14500438
diff --git a/tokenizer_config.json b/tokenizer_config.json
new file mode 100644
index 0000000000000000000000000000000000000000..c63345010532cae8bae2df6903ba2a811d5891d8
--- /dev/null
+++ b/tokenizer_config.json
@@ -0,0 +1 @@
+{"unk_token": "<unk>", "eos_token": "</s>", "bos_token": "<s>", "pad_token": "<pad>", "name_or_path": "bigscience/tokenizer", "special_tokens_map_file": null, "tokenizer_class": "BloomTokenizerFast"}
\ No newline at end of file