Upload ablation summaries

mbert_paper_metrics/docs/ablation_results.csv

@@ -78,6 +78,7 @@ result_ablation_mbert_paper,vsfc,mBERT,cls,42,eval_accuracy,0.932406822488945,0.
 result_ablation_mbert_paper,vsfc,mBERT,cls,43,eval_accuracy,0.9374605180037903,0.22002610564231873,0.22090962772324924,3.0,1583,/workspace/result_ablation_mbert_paper/vsfc/mBERT/cls/seed_43/all_results.json
 result_ablation_mbert_paper,vsfc,mBERT,cls,44,eval_accuracy,0.934301958307012,0.21943630278110504,0.232700464608786,3.0,1583,/workspace/result_ablation_mbert_paper/vsfc/mBERT/cls/seed_44/all_results.json
 result_ablation_mbert_paper,vsfc,mBERT,mean,42,eval_accuracy,0.9330385344283006,0.2111007124185562,0.2191808607194807,3.0,1583,/workspace/result_ablation_mbert_paper/vsfc/mBERT/mean/seed_42/all_results.json
+result_ablation_mbert_paper,vsfc,mBERT,mean,43,eval_accuracy,0.9349336702463676,0.21388660371303558,0.21989946765499516,3.0,1583,/workspace/result_ablation_mbert_paper/vsfc/mBERT/mean/seed_43/all_results.json
 result_ablation_mbert_paper,vsfc,mBERT,mean,44,eval_accuracy,0.932406822488945,0.21627604961395264,0.22658519144658443,3.0,1583,/workspace/result_ablation_mbert_paper/vsfc/mBERT/mean/seed_44/all_results.json
 result_ablation_mbert_paper,vsfc,mBERT,max,42,eval_accuracy,0.9336702463676564,0.2267763763666153,0.23415064422678558,3.0,1583,/workspace/result_ablation_mbert_paper/vsfc/mBERT/max/seed_42/all_results.json
 result_ablation_mbert_paper,vsfc,mBERT,max,43,eval_accuracy,0.9349336702463676,0.22572965919971466,0.23497871141055804,3.0,1583,/workspace/result_ablation_mbert_paper/vsfc/mBERT/max/seed_43/all_results.json

mbert_paper_metrics/docs/ablation_results_aggregate.csv

@@ -28,6 +28,6 @@ result_ablation_mbert_paper,vsfc,mBERT,cls,eval_accuracy,3,0.9347230995999157,0.
 result_ablation_mbert_paper,vsfc,mBERT,gated_multi_branch,eval_accuracy,3,0.9326173931353969,0.004741347757084709,0.9279848389134555,0.9374605180037903
 result_ablation_mbert_paper,vsfc,mBERT,hf_sequence_classifier,eval_accuracy,3,0.9317751105495894,0.0006317119393556281,0.9311433986102338,0.932406822488945
 result_ablation_mbert_paper,vsfc,mBERT,max,eval_accuracy,3,0.9340913876605601,0.0007294381164745449,0.9336702463676564,0.9349336702463676
-result_ablation_mbert_paper,vsfc,mBERT,mean,eval_accuracy,2,0.9327226784586229,0.0004466877960748697,0.932406822488945,0.9330385344283006
+result_ablation_mbert_paper,vsfc,mBERT,mean,eval_accuracy,3,0.9334596757212045,0.0013150132656134848,0.932406822488945,0.9349336702463676
 result_ablation_mbert_paper,vsfc,mBERT,mha_attention,eval_accuracy,3,0.9334596757212045,0.0007294381164746089,0.9330385344283006,0.934301958307012
 result_ablation_mbert_paper,vsfc,mBERT,multi_branch_average,eval_accuracy,1,0.9355653821857233,0.0,0.9355653821857233,0.9355653821857233

mbert_paper_metrics/docs/ablation_summary.md

@@ -34,7 +34,7 @@ Main metric is selected per task: CoLA uses Matthews correlation; MRPC/QQP/STSB
 | result_ablation_mbert_paper | vsfc | mBERT | gated_multi_branch | eval_accuracy | 3 | 0.9326 | 0.0047 | 0.9280 | 0.9375 |
 | result_ablation_mbert_paper | vsfc | mBERT | hf_sequence_classifier | eval_accuracy | 3 | 0.9318 | 0.0006 | 0.9311 | 0.9324 |
 | result_ablation_mbert_paper | vsfc | mBERT | max | eval_accuracy | 3 | 0.9341 | 0.0007 | 0.9337 | 0.9349 |
-| result_ablation_mbert_paper | vsfc | mBERT | mean | eval_accuracy | 2 | 0.9327 | 0.0004 | 0.9324 | 0.9330 |
+| result_ablation_mbert_paper | vsfc | mBERT | mean | eval_accuracy | 3 | 0.9335 | 0.0013 | 0.9324 | 0.9349 |
 | result_ablation_mbert_paper | vsfc | mBERT | mha_attention | eval_accuracy | 3 | 0.9335 | 0.0007 | 0.9330 | 0.9343 |
 | result_ablation_mbert_paper | vsfc | mBERT | multi_branch_average | eval_accuracy | 1 | 0.9356 | 0.0000 | 0.9356 | 0.9356 |
 
@@ -140,6 +140,7 @@ Main metric is selected per task: CoLA uses Matthews correlation; MRPC/QQP/STSB
 | result_ablation_mbert_paper | vsfc | mBERT | cls | 43.0000 | eval_accuracy | 0.9375 | 0.2200 | 0.2209 | 3.0000 | 1583 | /workspace/result_ablation_mbert_paper/vsfc/mBERT/cls/seed_43/all_results.json |
 | result_ablation_mbert_paper | vsfc | mBERT | cls | 44.0000 | eval_accuracy | 0.9343 | 0.2194 | 0.2327 | 3.0000 | 1583 | /workspace/result_ablation_mbert_paper/vsfc/mBERT/cls/seed_44/all_results.json |
 | result_ablation_mbert_paper | vsfc | mBERT | mean | 42.0000 | eval_accuracy | 0.9330 | 0.2111 | 0.2192 | 3.0000 | 1583 | /workspace/result_ablation_mbert_paper/vsfc/mBERT/mean/seed_42/all_results.json |
+| result_ablation_mbert_paper | vsfc | mBERT | mean | 43.0000 | eval_accuracy | 0.9349 | 0.2139 | 0.2199 | 3.0000 | 1583 | /workspace/result_ablation_mbert_paper/vsfc/mBERT/mean/seed_43/all_results.json |
 | result_ablation_mbert_paper | vsfc | mBERT | mean | 44.0000 | eval_accuracy | 0.9324 | 0.2163 | 0.2266 | 3.0000 | 1583 | /workspace/result_ablation_mbert_paper/vsfc/mBERT/mean/seed_44/all_results.json |
 | result_ablation_mbert_paper | vsfc | mBERT | max | 42.0000 | eval_accuracy | 0.9337 | 0.2268 | 0.2342 | 3.0000 | 1583 | /workspace/result_ablation_mbert_paper/vsfc/mBERT/max/seed_42/all_results.json |
 | result_ablation_mbert_paper | vsfc | mBERT | max | 43.0000 | eval_accuracy | 0.9349 | 0.2257 | 0.2350 | 3.0000 | 1583 | /workspace/result_ablation_mbert_paper/vsfc/mBERT/max/seed_43/all_results.json |