Hugging Face's logo

HavelockAI
/
bert-token-classifier

Token Classification
Transformers
Safetensors
English
modernbert
fill-mask
orality
linguistics
multi-label
custom_code
Model card Files
xet
Community
permutans commited on
Commit
3775141
·
verified ·
1 Parent(s): 86275ec

Upload folder using huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +153 -1
README.md CHANGED
@@ -131,6 +131,158 @@ Per-class F1 on test set (selected markers):
131
  | oral_audience_response | 1.000 | 1.000 | 1.000 | 4 |
132
  | literate_citation | 0.000 | 0.000 | 0.000 | 10 |
133
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
134
  **Macro F1 (all 145 labels):** 0.487
135
  **Weighted F1:** 0.645
136
  **Accuracy:** 66.5%
@@ -180,4 +332,4 @@ classifier.bias → randomly initialized
180
 
181
  ---
182
 
183
- *Model version: 668564aa • Trained: February 2026*
 
131
  | oral_audience_response | 1.000 | 1.000 | 1.000 | 4 |
132
  | literate_citation | 0.000 | 0.000 | 0.000 | 10 |
133
 
134
+ <details><summary>Click to show per-marker precision/recall/F1/support</summary>
135
+
136
+ ```
137
+ precision recall f1-score support
138
+
139
+ O 0.730 0.850 0.785 3496
140
+ B-literate_abstract_noun 0.500 0.071 0.125 14
141
+ B-literate_additive_formal 1.000 0.667 0.800 3
142
+ B-literate_agent_demoted 1.000 1.000 1.000 4
143
+ B-literate_agentless_passive 0.297 0.458 0.361 24
144
+ B-literate_aside 0.412 0.778 0.538 9
145
+ B-literate_categorical_statement 0.571 1.000 0.727 4
146
+ B-literate_causal_chain 0.667 0.667 0.667 3
147
+ B-literate_causal_explicit 0.467 0.636 0.538 11
148
+ B-literate_citation 0.000 0.000 0.000 10
149
+ B-literate_conceptual_metaphor 0.500 0.333 0.400 6
150
+ B-literate_concessive 1.000 1.000 1.000 2
151
+ B-literate_concessive_connector 0.800 0.800 0.800 5
152
+ B-literate_conditional 0.769 0.714 0.741 14
153
+ B-literate_contrastive 0.455 0.625 0.526 8
154
+ B-literate_definitional_move 1.000 1.000 1.000 1
155
+ B-literate_enumeration 0.429 1.000 0.600 3
156
+ B-literate_epistemic_hedge 0.357 0.417 0.385 24
157
+ B-literate_evidential 0.000 0.000 0.000 11
158
+ B-literate_footnote_reference 0.500 0.667 0.571 3
159
+ B-literate_institutional_subject 0.429 1.000 0.600 3
160
+ B-literate_list_structure 0.000 0.000 0.000 1
161
+ B-literate_metadiscourse 0.500 0.250 0.333 4
162
+ B-literate_methodological_framing 1.000 0.500 0.667 4
163
+ B-literate_nested_clauses 0.345 0.455 0.392 22
164
+ B-literate_nominalization 0.600 0.300 0.400 10
165
+ B-literate_objectifying_stance 0.800 1.000 0.889 4
166
+ B-literate_paradox 0.333 0.333 0.333 3
167
+ B-literate_probability 0.333 0.200 0.250 5
168
+ B-literate_qualified_assertion 0.000 0.000 0.000 5
169
+ B-literate_relative_chain 0.270 0.773 0.400 22
170
+ B-literate_technical_abbreviation 0.000 0.000 0.000 2
171
+ B-literate_technical_term 0.333 0.667 0.444 3
172
+ B-literate_temporal_embedding 1.000 0.500 0.667 4
173
+ B-literate_third_person_reference 0.500 0.333 0.400 3
174
+ B-oral_alliteration 0.000 0.000 0.000 3
175
+ B-oral_anaphora 0.188 0.400 0.255 15
176
+ B-oral_asyndeton 0.000 0.000 0.000 1
177
+ B-oral_audience_response 1.000 1.000 1.000 4
178
+ B-oral_binomial_expression 0.333 0.400 0.364 5
179
+ B-oral_conflict_frame 0.800 0.800 0.800 5
180
+ B-oral_discourse_formula 0.500 0.500 0.500 6
181
+ B-oral_dramatic_pause 1.000 0.500 0.667 2
182
+ B-oral_embodied_action 1.000 0.167 0.286 6
183
+ B-oral_epistrophe 0.500 0.333 0.400 3
184
+ B-oral_epithet 0.000 0.000 0.000 2
185
+ B-oral_everyday_example 0.750 1.000 0.857 3
186
+ B-oral_first_person 0.000 0.000 0.000 5
187
+ B-oral_imperative 0.571 0.571 0.571 14
188
+ B-oral_inclusive_we 0.500 0.586 0.540 29
189
+ B-oral_intensifier_doubling 1.000 0.667 0.800 3
190
+ B-oral_lexical_repetition 0.267 0.400 0.320 10
191
+ B-oral_named_individual 0.474 0.450 0.462 20
192
+ B-oral_parallelism 0.100 0.143 0.118 7
193
+ B-oral_phatic_check 1.000 1.000 1.000 1
194
+ B-oral_phatic_filler 0.429 0.600 0.500 5
195
+ B-oral_polysyndeton 0.200 0.100 0.133 10
196
+ B-oral_proverb 0.600 0.500 0.545 6
197
+ B-oral_refrain 1.000 1.000 1.000 1
198
+ B-oral_religious_formula 0.000 0.000 0.000 2
199
+ B-oral_rhetorical_question 0.125 0.500 0.200 2
200
+ B-oral_rhythm 0.714 0.833 0.769 6
201
+ B-oral_second_person 0.556 0.600 0.577 25
202
+ B-oral_self_correction 1.000 1.000 1.000 3
203
+ B-oral_sensory_detail 0.500 1.000 0.667 1
204
+ B-oral_simple_conjunction 0.000 0.000 0.000 3
205
+ B-oral_specific_place 0.333 0.667 0.444 3
206
+ B-oral_temporal_anchor 0.000 0.000 0.000 3
207
+ B-oral_tricolon 0.286 1.000 0.444 2
208
+ B-oral_us_them 0.667 0.667 0.667 3
209
+ B-oral_vocative 0.889 0.593 0.711 27
210
+ I-literate_abstract_noun 0.500 0.083 0.143 12
211
+ I-literate_additive_formal 0.000 0.000 0.000 6
212
+ I-literate_agent_demoted 0.583 0.933 0.718 15
213
+ I-literate_agentless_passive 0.518 0.414 0.460 70
214
+ I-literate_aside 0.700 0.618 0.656 102
215
+ I-literate_categorical_statement 0.692 0.391 0.500 23
216
+ I-literate_causal_chain 0.889 0.640 0.744 25
217
+ I-literate_causal_explicit 0.593 0.762 0.667 21
218
+ I-literate_citation 0.625 0.227 0.333 44
219
+ I-literate_conceptual_metaphor 0.857 0.400 0.545 15
220
+ I-literate_concessive 0.800 0.571 0.667 7
221
+ I-literate_concessive_connector 0.667 0.667 0.667 3
222
+ I-literate_conditional 0.784 0.351 0.485 114
223
+ I-literate_contrastive 0.857 0.400 0.545 15
224
+ I-literate_cross_reference 0.000 0.000 0.000 0
225
+ I-literate_definitional_move 1.000 1.000 1.000 5
226
+ I-literate_enumeration 0.680 0.436 0.531 39
227
+ I-literate_epistemic_hedge 0.425 0.386 0.405 44
228
+ I-literate_evidential 0.500 0.103 0.171 29
229
+ I-literate_footnote_reference 1.000 0.727 0.842 11
230
+ I-literate_institutional_subject 0.625 1.000 0.769 5
231
+ I-literate_list_structure 0.000 0.000 0.000 3
232
+ I-literate_metadiscourse 0.222 0.125 0.160 16
233
+ I-literate_methodological_framing 0.857 0.500 0.632 12
234
+ I-literate_nested_clauses 0.487 0.256 0.336 379
235
+ I-literate_nominalization 0.000 0.000 0.000 11
236
+ I-literate_objectifying_stance 0.900 0.692 0.783 13
237
+ I-literate_paradox 0.250 0.062 0.100 16
238
+ I-literate_probability 0.000 0.000 0.000 7
239
+ I-literate_qualified_assertion 0.000 0.000 0.000 21
240
+ I-literate_relative_chain 0.494 0.482 0.488 251
241
+ I-literate_technical_abbreviation 0.750 0.273 0.400 11
242
+ I-literate_technical_term 0.167 0.273 0.207 11
243
+ I-literate_temporal_embedding 1.000 0.600 0.750 50
244
+ I-literate_third_person_reference 0.714 0.833 0.769 6
245
+ I-oral_alliteration 1.000 0.222 0.364 9
246
+ I-oral_anaphora 0.140 0.070 0.093 100
247
+ I-oral_asyndeton 0.417 0.714 0.526 7
248
+ I-oral_audience_response 0.905 0.905 0.905 21
249
+ I-oral_binomial_expression 0.529 0.818 0.643 11
250
+ I-oral_conflict_frame 0.833 0.714 0.769 7
251
+ I-oral_discourse_formula 0.625 0.833 0.714 6
252
+ I-oral_dramatic_pause 0.500 0.500 0.500 4
253
+ I-oral_embodied_action 1.000 0.062 0.118 16
254
+ I-oral_epistrophe 0.000 0.000 0.000 3
255
+ I-oral_epithet 0.200 0.200 0.200 5
256
+ I-oral_everyday_example 1.000 0.950 0.974 20
257
+ I-oral_first_person 0.000 0.000 0.000 2
258
+ I-oral_imperative 0.692 0.333 0.450 27
259
+ I-oral_inclusive_we 0.849 0.900 0.874 50
260
+ I-oral_intensifier_doubling 0.800 1.000 0.889 4
261
+ I-oral_lexical_repetition 0.241 0.341 0.283 41
262
+ I-oral_named_individual 0.689 0.646 0.667 48
263
+ I-oral_parallelism 0.518 0.319 0.394 135
264
+ I-oral_phatic_check 1.000 1.000 1.000 3
265
+ I-oral_phatic_filler 0.667 0.400 0.500 5
266
+ I-oral_polysyndeton 1.000 0.268 0.423 82
267
+ I-oral_proverb 0.955 0.568 0.712 37
268
+ I-oral_refrain 1.000 1.000 1.000 4
269
+ I-oral_religious_formula 0.000 0.000 0.000 16
270
+ I-oral_rhetorical_question 0.312 0.333 0.323 15
271
+ I-oral_rhythm 0.979 0.610 0.752 77
272
+ I-oral_second_person 0.143 0.143 0.143 7
273
+ I-oral_self_correction 0.857 0.900 0.878 20
274
+ I-oral_sensory_detail 0.833 1.000 0.909 5
275
+ I-oral_simple_conjunction 0.714 1.000 0.833 5
276
+ I-oral_specific_place 0.714 0.625 0.667 8
277
+ I-oral_temporal_anchor 0.059 0.100 0.074 10
278
+ I-oral_tricolon 0.388 0.929 0.547 28
279
+ I-oral_us_them 0.800 0.889 0.842 9
280
+ I-oral_vocative 0.971 0.723 0.829 47
281
+ ```
282
+
283
+ </details>
284
+
285
+
286
  **Macro F1 (all 145 labels):** 0.487
287
  **Weighted F1:** 0.645
288
  **Accuracy:** 66.5%
 
332
 
333
  ---
334
 
335
+ *Model version: 668564aa • Trained: February 2026*