BayanDuygu commited on
Commit
33ac258
·
verified ·
1 Parent(s): a18e7df

added bench results

Browse files
Files changed (1) hide show
  1. README.md +46 -3
README.md CHANGED
@@ -85,12 +85,16 @@ configs:
85
  - split: validation
86
  path: IMST/dev.jsonl
87
  ---
88
- ---
89
 
90
 
91
  <img src="https://raw.githubusercontent.com/turkish-nlp-suite/.github/main/profile/MediTurcaTextlogo.png" width="30%" height="30%">
92
 
93
- # Turkkish Treebank Benchmarking
 
 
 
 
94
 
95
  | Dataset | train size | dev size | test size|
96
  |---|---|---|---|
@@ -103,8 +107,47 @@ configs:
103
 
104
  A typical instance from the dataset looks like:
105
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
106
 
107
 
108
 
109
  ## Acknowledgments
110
- This research was supported with Cloud TPUs from Google's TPU Research Cloud (TRC).
 
85
  - split: validation
86
  path: IMST/dev.jsonl
87
  ---
88
+
89
 
90
 
91
  <img src="https://raw.githubusercontent.com/turkish-nlp-suite/.github/main/profile/MediTurcaTextlogo.png" width="30%" height="30%">
92
 
93
+ # Turkish Treebank Benchmarking
94
+ This is the repo for Turkish treebank benchmarking, namely evaluating Tranformer models on POS-DEP-MORPH task.
95
+ For the data, we used two datasets, [IMST]() and [BOUN](). We converted conllu format to json lines for being compatible to HF dataset formats.
96
+
97
+ Here are treebank sizes at a glance:
98
 
99
  | Dataset | train size | dev size | test size|
100
  |---|---|---|---|
 
107
 
108
  A typical instance from the dataset looks like:
109
 
110
+ ```
111
+ ```
112
+
113
+ ## Benchmarking
114
+
115
+
116
+
117
+ Here are the benchmarking results for BERTurk:
118
+
119
+ ### Test results (BOUN vs IMST)
120
+
121
+ | Metric | BOUN | IMST |
122
+ |---|---:|---:|
123
+ | pos_acc | 0.9263 | 0.9377 |
124
+ | uas | 0.8151 | 0.7680 |
125
+ | las | 0.7459 | 0.6960 |
126
+ | morph_Abbr_acc | 0.4657 | 0.6705 |
127
+ | morph_Aspect_acc | 0.1141 | 0.1152 |
128
+ | morph_Case_acc | 0.1196 | 0.0586 |
129
+ | morph_Echo_acc | 0.4261 | 0.4875 |
130
+ | morph_Evident_acc | 0.3072 | 0.3953 |
131
+ | morph_Mood_acc | 0.0654 | 0.0651 |
132
+ | morph_NumType_acc | 0.2694 | 0.2991 |
133
+ | morph_Number_acc | 0.3986 | 0.4782 |
134
+ | morph_Number[psor]_acc | 0.4348 | 0.2333 |
135
+ | morph_Person_acc | 0.4021 | 0.4726 |
136
+ | morph_Person[psor]_acc | 0.2490 | 0.0671 |
137
+ | morph_Polarity_acc | 0.3350 | 0.1674 |
138
+ | morph_PronType_acc | 0.1535 | 0.2680 |
139
+ | morph_Reflex_acc | 0.5620 | 0.7051 |
140
+ | morph_Tense_acc | 0.2149 | 0.1241 |
141
+ | morph_Typo_acc | 0.5081 | — |
142
+ | morph_VerbForm_acc | 0.4912 | 0.2364 |
143
+ | morph_Voice_acc | 0.0201 | 0.2602 |
144
+ | morph_Polite_acc | — | 0.1436 |
145
+ | morph_micro_acc | 0.3076 | 0.2915 |
146
+
147
+ Notes:
148
+ - `—` means that metric wasn’t present in that dataset’s reported results (e.g., `morph_Typo_acc` only in BOUN; `morph_Polite_acc` only in IMST).
149
 
150
 
151
 
152
  ## Acknowledgments
153
+ This research was supported with Cloud TPUs from Google's TPU Research Cloud (TRC), like most of our projects. Many thanks to TRC team once again.