Model_name
stringclasses
16 values
Train_size
int64
50.8k
50.8k
Test_size
int64
12.7k
12.7k
arg
dict
lora
listlengths
1
9
Parameters
int64
110M
1.85B
Trainable_parameters
int64
9.27k
1.11B
r
int64
4
1.02k
Memory Allocation
stringlengths
5
7
Training Time
stringlengths
5
7
Performance
dict
albert/albert-xxlarge-v2
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "embedding_hidden_mapping_in", "ffn", "ffn_output", "key", "pooler", "query", "value" ]
223,391,258
742,413
8
2587.94
3250.57
{ "accuracy": 0.8722731583939298, "f1_macro": 0.862150429106117, "f1_weighted": 0.8723422017705071, "precision": 0.8658039080136887, "recall": 0.8593864161213076 }
facebook/opt-350m
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "fc1", "fc2", "k_proj", "out_proj", "project_in", "project_out", "q_proj", "score", "v_proj" ]
345,463,808
14,260,736
32
3288.99
1654.49
{ "accuracy": 0.8845241858994626, "f1_macro": 0.8791726339899201, "f1_weighted": 0.8846846511262262, "precision": 0.8798342880590895, "recall": 0.8787539901529715 }
FacebookAI/xlm-roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
568,044,570
8,140,813
16
5030.66
2209.58
{ "accuracy": 0.889661713563073, "f1_macro": 0.8848712853323799, "f1_weighted": 0.8898577947681364, "precision": 0.8870786826880624, "recall": 0.8830419063481568 }
FacebookAI/roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
370,591,770
15,218,701
32
3508.76
1661.64
{ "accuracy": 0.89116345241859, "f1_macro": 0.885951353706553, "f1_weighted": 0.8914312375722376, "precision": 0.886857455081974, "recall": 0.8853286180328447 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
422,033,434
15,807,501
32
3792.82
1905.48
{ "accuracy": 0.8889503635788808, "f1_macro": 0.8817753174512746, "f1_weighted": 0.8891663020744098, "precision": 0.8831525856714956, "recall": 0.8807908131673957 }
google-t5/t5-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "k", "o", "out_proj", "q", "v", "wi", "wo" ]
756,081,885
17,350,864
16
2361.3
2706.95
{ "accuracy": 0.82042364843503, "f1_macro": 0.7833049300677861, "f1_weighted": 0.8144741005682286, "precision": 0.8180956093532021, "recall": 0.7822013600956367 }
Qwen/Qwen3-Reranker-0.6B
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "down_proj", "gate_proj", "k_proj", "o_proj", "q_proj", "score", "up_proj", "v_proj" ]
636,173,312
40,383,488
64
4360.82
1491.26
{ "accuracy": 0.8875276636104964, "f1_macro": 0.881302318996792, "f1_weighted": 0.8876169297855617, "precision": 0.8833784383227032, "recall": 0.879686542527803 }
facebook/opt-350m
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "fc1", "fc2", "k_proj", "out_proj", "project_in", "project_out", "q_proj", "score", "v_proj" ]
359,717,888
28,514,816
64
3501.25
1725.59
{ "accuracy": 0.8891084413531457, "f1_macro": 0.884251353876662, "f1_weighted": 0.8893208862840672, "precision": 0.8852089687922547, "recall": 0.8835820879208923 }
FacebookAI/roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
384,747,546
29,374,477
64
3717.72
1734.12
{ "accuracy": 0.8947992412266835, "f1_macro": 0.8899736041154382, "f1_weighted": 0.8950306912490653, "precision": 0.8906489322663366, "recall": 0.8895635950169705 }
FacebookAI/xlm-roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
575,122,458
15,218,701
32
5144.48
2235.28
{ "accuracy": 0.8925071134998419, "f1_macro": 0.8879024011502923, "f1_weighted": 0.8926429337051776, "precision": 0.8897393465596863, "recall": 0.8863225423457599 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
437,827,610
31,601,677
64
4058.01
1973.96
{ "accuracy": 0.8944040467910211, "f1_macro": 0.888322020211215, "f1_weighted": 0.8946660494482694, "precision": 0.8889590332743614, "recall": 0.8879002544876452 }
Qwen/Qwen3-Reranker-0.6B
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "down_proj", "gate_proj", "k_proj", "o_proj", "q_proj", "score", "up_proj", "v_proj" ]
676,543,488
80,753,664
128
5036.09
1592.53
{ "accuracy": 0.8925861523869744, "f1_macro": 0.8868377941363138, "f1_weighted": 0.8926828164233866, "precision": 0.8886165937671825, "recall": 0.885368684957202 }
albert/albert-xxlarge-v2
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "embedding_hidden_mapping_in", "ffn", "ffn_output", "key", "pooler", "query", "value" ]
224,080,410
1,431,565
16
2593.0
3261.23
{ "accuracy": 0.8821530192854885, "f1_macro": 0.874794659895967, "f1_weighted": 0.8823534689722649, "precision": 0.8770079907164144, "recall": 0.8731264865543866 }
facebook/opt-350m
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "fc1", "fc2", "k_proj", "out_proj", "project_in", "project_out", "q_proj", "score", "v_proj" ]
388,226,048
57,022,976
128
3932.16
1872.85
{ "accuracy": 0.8920328801770471, "f1_macro": 0.8868732981442506, "f1_weighted": 0.8922332162153938, "precision": 0.8877692584168926, "recall": 0.8862711257051747 }
google-t5/t5-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "k", "o", "out_proj", "q", "v", "wi", "wo" ]
773,432,749
34,701,728
32
2475.04
2700.75
{ "accuracy": 0.8593107809042049, "f1_macro": 0.8440172091473165, "f1_weighted": 0.8585287294204809, "precision": 0.8532288071809823, "recall": 0.8396354174184749 }
FacebookAI/roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
413,059,098
57,686,029
128
4149.31
1883.44
{ "accuracy": 0.8980398355991147, "f1_macro": 0.8933785290970504, "f1_weighted": 0.8982180255485378, "precision": 0.8943152546471684, "recall": 0.8926576739287886 }
FacebookAI/xlm-roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
589,278,234
29,374,477
64
5353.26
2310.3
{ "accuracy": 0.8963009800822005, "f1_macro": 0.8929057205031824, "f1_weighted": 0.8964683685160553, "precision": 0.8940441682799969, "recall": 0.8919755114442118 }
Qwen/Qwen3-Reranker-0.6B
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "down_proj", "gate_proj", "k_proj", "o_proj", "q_proj", "score", "up_proj", "v_proj" ]
757,283,840
161,494,016
256
6152.73
1764.92
{ "accuracy": 0.8938507745810939, "f1_macro": 0.8888681143587468, "f1_weighted": 0.8940071272374931, "precision": 0.8908125929555937, "recall": 0.8872898702169905 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
469,415,962
63,190,029
128
4575.04
2116.9
{ "accuracy": 0.9004110022130888, "f1_macro": 0.8957841401977072, "f1_weighted": 0.9005870758571973, "precision": 0.8966880236436467, "recall": 0.8950678952185568 }
facebook/opt-350m
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "fc1", "fc2", "k_proj", "out_proj", "project_in", "project_out", "q_proj", "score", "v_proj" ]
445,242,368
114,039,296
256
4697.46
2132.45
{ "accuracy": 0.8953525134366108, "f1_macro": 0.8910942257195085, "f1_weighted": 0.8955565005279739, "precision": 0.8922545500690322, "recall": 0.8901897326636318 }
FacebookAI/roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
469,682,202
114,309,133
256
4939.87
2136.19
{ "accuracy": 0.8983559911476446, "f1_macro": 0.8934100039534362, "f1_weighted": 0.8985384020635359, "precision": 0.8942080932451125, "recall": 0.892813089993333 }
google-t5/t5-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "k", "o", "out_proj", "q", "v", "wi", "wo" ]
808,134,477
69,403,456
64
2703.21
2738.66
{ "accuracy": 0.8797818526715144, "f1_macro": 0.8699565041745851, "f1_weighted": 0.8800662980988457, "precision": 0.8729282824509633, "recall": 0.8679609354136181 }
albert/albert-xxlarge-v2
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "embedding_hidden_mapping_in", "ffn", "ffn_output", "key", "pooler", "query", "value" ]
225,458,714
2,809,869
32
2604.14
3278.95
{ "accuracy": 0.8866582358520392, "f1_macro": 0.8805361394786488, "f1_weighted": 0.8868472161794123, "precision": 0.8821746780557068, "recall": 0.8792168958629811 }
Qwen/Qwen3-Reranker-0.6B
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "down_proj", "gate_proj", "k_proj", "o_proj", "q_proj", "score", "up_proj", "v_proj" ]
918,764,544
322,974,720
512
8024.28
2170.38
{ "accuracy": 0.8966961745178628, "f1_macro": 0.8918835728777867, "f1_weighted": 0.896871052710265, "precision": 0.8933411271456468, "recall": 0.8907348982342026 }
FacebookAI/xlm-roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
617,589,786
57,686,029
128
5781.26
2466.76
{ "accuracy": 0.8970123300663927, "f1_macro": 0.8932253980605369, "f1_weighted": 0.8972005164623214, "precision": 0.8944153044334617, "recall": 0.8922802938128591 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
532,592,666
126,366,733
256
5391.17
2386.94
{ "accuracy": 0.9015965855200759, "f1_macro": 0.8973526640185696, "f1_weighted": 0.9018252959007775, "precision": 0.8984521946549535, "recall": 0.8965587386239398 }
facebook/opt-350m
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "fc1", "fc2", "k_proj", "out_proj", "project_in", "project_out", "q_proj", "score", "v_proj" ]
559,275,008
228,071,936
512
6303.52
2681.3
{ "accuracy": 0.896063863420803, "f1_macro": 0.8914847709874795, "f1_weighted": 0.8962464034480218, "precision": 0.8925647716667204, "recall": 0.8907047374890307 }
FacebookAI/roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
582,928,410
227,555,341
512
6516.59
2688.97
{ "accuracy": 0.902466013278533, "f1_macro": 0.8982951782674592, "f1_weighted": 0.9026722067520413, "precision": 0.898378896740835, "recall": 0.8984416803597521 }
google-t5/t5-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "k", "o", "out_proj", "q", "v", "wi", "wo" ]
877,537,933
138,806,912
128
3194.29
2737.1
{ "accuracy": 0.8878438191590262, "f1_macro": 0.8807570808340444, "f1_weighted": 0.8882097376153891, "precision": 0.8827566138741335, "recall": 0.8794871258073358 }
FacebookAI/xlm-roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
674,212,890
114,309,133
256
6556.64
2727.88
{ "accuracy": 0.8993044577932343, "f1_macro": 0.8953449229210466, "f1_weighted": 0.8996268958775625, "precision": 0.8959634381136734, "recall": 0.8951316615673002 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
658,946,074
252,720,141
512
7180.6
2927.03
{ "accuracy": 0.9049162187796396, "f1_macro": 0.901107386190174, "f1_weighted": 0.9051770409666084, "precision": 0.901329006868039, "recall": 0.9011179531520016 }
albert/albert-xxlarge-v2
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "embedding_hidden_mapping_in", "ffn", "ffn_output", "key", "pooler", "query", "value" ]
228,215,322
5,566,477
64
2627.37
3338.35
{ "accuracy": 0.8893455580145432, "f1_macro": 0.8831911868614246, "f1_weighted": 0.8895443767065669, "precision": 0.8843768003085015, "recall": 0.8822588568756525 }
google-t5/t5-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "k", "o", "out_proj", "q", "v", "wi", "wo" ]
1,016,344,845
277,613,824
256
4094.39
2784.5
{ "accuracy": 0.8972494467277901, "f1_macro": 0.8914586129914629, "f1_weighted": 0.8975192403879136, "precision": 0.8923177953104979, "recall": 0.8909761994005975 }
FacebookAI/xlm-roberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
787,459,098
227,555,341
512
8153.74
3294.49
{ "accuracy": 0.9019127410686057, "f1_macro": 0.8987907740359878, "f1_weighted": 0.9020500820267734, "precision": 0.8993803181920761, "recall": 0.8983952376220622 }
albert/albert-xxlarge-v2
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "embedding_hidden_mapping_in", "ffn", "ffn_output", "key", "pooler", "query", "value" ]
233,728,538
11,079,693
128
2669.5
3444.4
{ "accuracy": 0.8938507745810939, "f1_macro": 0.888144143643366, "f1_weighted": 0.8940975232304162, "precision": 0.8890617363540988, "recall": 0.8875081136269234 }
google-t5/t5-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "k", "o", "out_proj", "q", "v", "wi", "wo" ]
1,293,958,669
555,227,648
512
6137.83
2885.91
{ "accuracy": 0.9001738855516914, "f1_macro": 0.894299046872226, "f1_weighted": 0.900442163738966, "precision": 0.8947516066386669, "recall": 0.8941409421046498 }
albert/albert-xxlarge-v2
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "embedding_hidden_mapping_in", "ffn", "ffn_output", "key", "pooler", "query", "value" ]
244,754,970
22,106,125
256
2765.5
3658.44
{ "accuracy": 0.896380018969333, "f1_macro": 0.8904405976694844, "f1_weighted": 0.8965934269190462, "precision": 0.8909890016674086, "recall": 0.8901285257305411 }
google-t5/t5-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "k", "o", "out_proj", "q", "v", "wi", "wo" ]
1,849,186,317
1,110,455,296
1,024
9704.44
3808.17
{ "accuracy": 0.9046000632311098, "f1_macro": 0.8996896264023401, "f1_weighted": 0.9048083267959188, "precision": 0.9001987207094116, "recall": 0.8994839753259615 }
albert/albert-xxlarge-v2
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "embedding_hidden_mapping_in", "ffn", "ffn_output", "key", "pooler", "query", "value" ]
266,807,834
44,158,989
512
2948.18
4128.85
{ "accuracy": 0.8989883022447044, "f1_macro": 0.8932787242488319, "f1_weighted": 0.899180989072891, "precision": 0.8939046353162913, "recall": 0.8929198305459212 }
FacebookAI/roberta-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
125,919,770
1,264,141
4
1391.99
649.89
{ "accuracy": 0.8484824533670566, "f1_macro": 0.8358528600947209, "f1_weighted": 0.8484053097256183, "precision": 0.8414743158781716, "recall": 0.8322497328689692 }
google-bert/bert-base-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
110,171,930
679,693
4
1260.55
604.22
{ "accuracy": 0.6776003793866582, "f1_macro": 0.557677468110299, "f1_weighted": 0.6360221330998432, "precision": 0.5554900480313727, "recall": 0.5934898807250314 }
FacebookAI/roberta-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
126,583,322
1,927,693
8
1392.71
647.49
{ "accuracy": 0.862867530825166, "f1_macro": 0.8530823288347613, "f1_weighted": 0.8629893081564594, "precision": 0.8579622374628227, "recall": 0.8495085850791969 }
google-bert/bert-base-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
110,841,626
1,349,389
8
1260.41
604.98
{ "accuracy": 0.7556907998735378, "f1_macro": 0.6913816116702359, "f1_weighted": 0.7370133650405578, "precision": 0.7621957139881789, "recall": 0.6991076757647363 }
FacebookAI/roberta-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
127,910,426
3,254,797
16
1402.33
659.61
{ "accuracy": 0.8700600695542207, "f1_macro": 0.8617076949119702, "f1_weighted": 0.8702660794549991, "precision": 0.8650290907522602, "recall": 0.8590312471174008 }
facebook/bart-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "k_proj", "out_proj", "q_proj", "v_proj" ]
409,529,409
2,175,028
4
3931.55
1994.23
{ "accuracy": 0.837100853619981, "f1_macro": 0.819320002150968, "f1_weighted": 0.8365214467882119, "precision": 0.829413114853224, "recall": 0.8146486251583185 }
google-bert/bert-base-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
112,181,018
2,688,781
16
1271.48
614.52
{ "accuracy": 0.8084097375908947, "f1_macro": 0.7752610196307308, "f1_weighted": 0.8023462795387738, "precision": 0.8061572719439902, "recall": 0.7713207183347648 }
google-bert/bert-large-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
336,946,202
1,790,989
4
3144.02
1682.06
{ "accuracy": 0.7811413215301929, "f1_macro": 0.7194840912752345, "f1_weighted": 0.7656902509079089, "precision": 0.7873359402685902, "recall": 0.7243830682157009 }
FacebookAI/roberta-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
130,564,634
5,909,005
32
1455.37
663.66
{ "accuracy": 0.8754347138792286, "f1_macro": 0.8678486993252565, "f1_weighted": 0.8756572384400854, "precision": 0.8698494102676955, "recall": 0.8662188276261766 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
408,213,530
1,987,597
4
3615.83
1929.23
{ "accuracy": 0.8498261144483086, "f1_macro": 0.8320999499486269, "f1_weighted": 0.8489416106176354, "precision": 0.8392730685236397, "recall": 0.8282218416505667 }
google-bert/bert-base-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
114,859,802
5,367,565
32
1324.87
617.48
{ "accuracy": 0.8453999367688902, "f1_macro": 0.832204269972286, "f1_weighted": 0.8452577560841472, "precision": 0.8415721920329958, "recall": 0.8270353900732399 }
FacebookAI/roberta-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
135,873,050
11,217,421
64
1516.47
676.66
{ "accuracy": 0.8810464748656339, "f1_macro": 0.8736485406190957, "f1_weighted": 0.8812054216381876, "precision": 0.8750473261159748, "recall": 0.8724947273133411 }
google-bert/bert-base-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
120,217,370
10,725,133
64
1386.91
630.21
{ "accuracy": 0.8607334808725893, "f1_macro": 0.8507968201349172, "f1_weighted": 0.8608113855611761, "precision": 0.8563174244998447, "recall": 0.8472312217095838 }
google-bert/bert-large-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
338,723,866
3,568,653
8
3164.27
1685.52
{ "accuracy": 0.8390768257982928, "f1_macro": 0.8215203388033808, "f1_weighted": 0.8378995988408484, "precision": 0.8353637674216158, "recall": 0.8142267320449833 }
FacebookAI/roberta-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
146,489,882
21,834,253
128
1560.5
702.94
{ "accuracy": 0.8858678469807145, "f1_macro": 0.8792843892984809, "f1_weighted": 0.8859717818352927, "precision": 0.8805172243986237, "recall": 0.878249432241028 }
google-bert/bert-base-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
130,932,506
21,440,269
128
1437.65
655.95
{ "accuracy": 0.8727473917167247, "f1_macro": 0.8657257994511084, "f1_weighted": 0.8728823790503637, "precision": 0.8683363006590009, "recall": 0.863745947792565 }
facebook/bart-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "k_proj", "out_proj", "q_proj", "v_proj" ]
411,704,437
4,350,056
8
3959.44
1961.57
{ "accuracy": 0.8593107809042049, "f1_macro": 0.8451630147794547, "f1_weighted": 0.8592830424157094, "precision": 0.8497632268604688, "recall": 0.8425814401047079 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
410,187,802
3,961,869
8
3634.97
1931.87
{ "accuracy": 0.8739329750237117, "f1_macro": 0.8630749222424812, "f1_weighted": 0.8739608067544079, "precision": 0.8660347921507289, "recall": 0.8612344922001133 }
FacebookAI/roberta-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
167,723,546
43,067,917
256
1870.04
790.67
{ "accuracy": 0.8887132469174834, "f1_macro": 0.8823514130946098, "f1_weighted": 0.8887802619711761, "precision": 0.8838996386590717, "recall": 0.880989032450131 }
google-bert/bert-base-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
152,362,778
42,870,541
256
1746.14
738.67
{ "accuracy": 0.8804141637685742, "f1_macro": 0.8745999258330277, "f1_weighted": 0.8805422577690016, "precision": 0.8765720714268965, "recall": 0.8731069954938984 }
google-bert/bert-large-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
342,279,194
7,123,981
16
3213.34
1700.35
{ "accuracy": 0.8574928865001581, "f1_macro": 0.846890119031086, "f1_weighted": 0.85781495973405, "precision": 0.8518450581081903, "recall": 0.8434738422594921 }
FacebookAI/roberta-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "key", "out_proj", "query", "value" ]
210,190,874
85,535,245
512
2378.39
962.47
{ "accuracy": 0.8916376857413848, "f1_macro": 0.8857860565959809, "f1_weighted": 0.8917251879556186, "precision": 0.8869669092700435, "recall": 0.8847594241994637 }
google-bert/bert-base-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
195,223,322
85,731,085
512
2274.07
911.63
{ "accuracy": 0.88444514701233, "f1_macro": 0.8791416361373022, "f1_weighted": 0.8845230747528955, "precision": 0.8807529967164646, "recall": 0.8778948114704137 }
facebook/bart-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "k_proj", "out_proj", "q_proj", "v_proj" ]
416,054,493
8,700,112
16
4012.11
1986.69
{ "accuracy": 0.8701391084413531, "f1_macro": 0.8590196347959178, "f1_weighted": 0.8702395584157556, "precision": 0.8621420732579376, "recall": 0.857157647104303 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
414,136,346
7,910,413
16
3683.47
1944.31
{ "accuracy": 0.8834966803667405, "f1_macro": 0.8752151288172615, "f1_weighted": 0.8836837903837133, "precision": 0.8774345344074861, "recall": 0.8735503023918401 }
google-bert/bert-large-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
349,389,850
14,234,637
32
3328.52
1726.62
{ "accuracy": 0.8693487195700285, "f1_macro": 0.8615475713243064, "f1_weighted": 0.8698061074222874, "precision": 0.864004754302695, "recall": 0.8599117843638718 }
facebook/bart-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "k_proj", "out_proj", "q_proj", "v_proj" ]
424,754,605
17,400,224
32
4142.27
2025.66
{ "accuracy": 0.8789914638001897, "f1_macro": 0.8705082136246272, "f1_weighted": 0.8792583179592576, "precision": 0.8734569164370962, "recall": 0.8685444688611441 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
422,033,434
15,807,501
32
3792.82
1973.37
{ "accuracy": 0.8881599747075561, "f1_macro": 0.880613244415908, "f1_weighted": 0.8883373228402671, "precision": 0.8825283842549392, "recall": 0.8792305171377859 }
google-bert/bert-large-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
363,611,162
28,455,949
64
3536.97
1803.29
{ "accuracy": 0.8771735693961429, "f1_macro": 0.8699330967814549, "f1_weighted": 0.8774191862467641, "precision": 0.8714080281561766, "recall": 0.8688668759782217 }
facebook/bart-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "k_proj", "out_proj", "q_proj", "v_proj" ]
442,154,829
34,800,448
64
4405.16
2105.68
{ "accuracy": 0.8857097692064496, "f1_macro": 0.8789842826741208, "f1_weighted": 0.8860069708455789, "precision": 0.8800323440010615, "recall": 0.8783776772722253 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
437,827,610
31,601,677
64
4058.01
2047.57
{ "accuracy": 0.8944830856781536, "f1_macro": 0.8888925420693033, "f1_weighted": 0.8946779369114773, "precision": 0.8899960829353635, "recall": 0.8880635347919621 }
google-bert/bert-large-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
392,053,786
56,898,573
128
3975.7
1952.99
{ "accuracy": 0.8834966803667405, "f1_macro": 0.8783752303231847, "f1_weighted": 0.88368914760437, "precision": 0.8794481315532393, "recall": 0.8776567718517531 }
facebook/bart-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "k_proj", "out_proj", "q_proj", "v_proj" ]
476,955,277
69,600,896
128
4967.39
2298.46
{ "accuracy": 0.8925071134998419, "f1_macro": 0.8874563535314081, "f1_weighted": 0.8927586350868724, "precision": 0.8879377177918474, "recall": 0.8873768776967422 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
469,415,962
63,190,029
128
4573.34
2199.88
{ "accuracy": 0.9004900411002214, "f1_macro": 0.8957679639409909, "f1_weighted": 0.9006955885096265, "precision": 0.8967474874901942, "recall": 0.8950164503701811 }
google-bert/bert-large-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
448,939,034
113,783,821
256
4750.65
2213.83
{ "accuracy": 0.8874486247233639, "f1_macro": 0.8823049883630817, "f1_weighted": 0.8876614939241605, "precision": 0.8829447914365554, "recall": 0.8819320015084025 }
facebook/bart-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "k_proj", "out_proj", "q_proj", "v_proj" ]
546,556,173
139,201,792
256
5891.24
2603.79
{ "accuracy": 0.8970123300663927, "f1_macro": 0.8923001378224603, "f1_weighted": 0.8971946194525198, "precision": 0.8927613365832312, "recall": 0.8921285652029669 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
532,592,666
126,366,733
256
5391.17
2487.22
{ "accuracy": 0.9027031299399304, "f1_macro": 0.8985249989118346, "f1_weighted": 0.9029153635967133, "precision": 0.8995863769514236, "recall": 0.8977173916793645 }
google-bert/bert-large-uncased
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "key", "query", "value" ]
562,709,530
227,554,317
512
6351.86
2787.0
{ "accuracy": 0.8914005690799873, "f1_macro": 0.8864933880817142, "f1_weighted": 0.8915764179384021, "precision": 0.8871029944503552, "recall": 0.8861094670610424 }
microsoft/deberta-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "classifier", "dense", "in_proj", "pos_proj", "pos_q_proj" ]
658,946,074
252,720,141
512
7180.6
3068.94
{ "accuracy": 0.9066550742965539, "f1_macro": 0.9027440214865444, "f1_weighted": 0.9068737680350811, "precision": 0.9032737095991187, "recall": 0.9024190025341322 }
facebook/bart-large
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "k_proj", "out_proj", "q_proj", "v_proj" ]
685,757,965
278,403,584
512
7867.37
3278.36
{ "accuracy": 0.9001738855516914, "f1_macro": 0.8959184453049027, "f1_weighted": 0.9003404184427293, "precision": 0.8963869391852941, "recall": 0.8956855215198263 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "out_proj" ]
408,646,721
1,290,292
4
953.04
781.26
{ "accuracy": 0.845004742333228, "f1_macro": 0.8291413894360475, "f1_weighted": 0.8448352931372985, "precision": 0.8350869160708587, "recall": 0.8256532089018532 }
google/flan-t5-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "out_proj" ]
223,513,409
9,268
4
1828.27
794.45
{ "accuracy": 0.12622510275055326, "f1_macro": 0.061089496032435624, "f1_weighted": 0.07495071441669991, "precision": 0.08000500651141326, "recall": 0.10210739144307183 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "out_proj" ]
409,937,013
2,580,584
8
944.26
789.93
{ "accuracy": 0.8638950363578881, "f1_macro": 0.8525358873374469, "f1_weighted": 0.8642250368019567, "precision": 0.8567960191676632, "recall": 0.8499438528166792 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "out_proj" ]
412,517,597
5,161,168
16
960.83
789.47
{ "accuracy": 0.8725893139424596, "f1_macro": 0.8632568652072942, "f1_weighted": 0.8728727834211805, "precision": 0.8662148817287414, "recall": 0.8611576784632347 }
google/flan-t5-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k", "o", "q", "v", "wi_0", "wi_1", "wo" ]
225,209,153
1,705,012
4
2649.39
1710.46
{ "accuracy": 0.19680682895984825, "f1_macro": 0.09519672524418793, "f1_weighted": 0.11695216801550407, "precision": 0.11552054112946236, "recall": 0.15800505790358427 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "out_proj" ]
417,678,765
10,322,336
32
993.96
786.86
{ "accuracy": 0.8827853303825483, "f1_macro": 0.8756500209958136, "f1_weighted": 0.8831353714515359, "precision": 0.8771788767815625, "recall": 0.8745545422664613 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "out_proj" ]
428,001,101
20,644,672
64
1060.35
794.72
{ "accuracy": 0.8885551691432184, "f1_macro": 0.882663097395477, "f1_weighted": 0.8888437665388873, "precision": 0.8832717599594526, "recall": 0.8823910207284164 }
google/flan-t5-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "wi_0", "dense", "v", "wi_1", "k", "out_proj", "q", "wo", "o" ]
225,209,153
1,705,012
4
2640.87
1719.28
{ "accuracy": 0.19008852355358838, "f1_macro": 0.09162888412305616, "f1_weighted": 0.11225337817161031, "precision": 0.10799035490498364, "recall": 0.15256261921842296 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "out_proj" ]
448,645,773
41,289,344
128
1191.16
798.64
{ "accuracy": 0.8938507745810939, "f1_macro": 0.8886604857922034, "f1_weighted": 0.8940994763136924, "precision": 0.8894674968455903, "recall": 0.8881550093984057 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "fc1", "fc2", "out_proj" ]
489,935,117
82,578,688
256
1475.45
847.16
{ "accuracy": 0.8988302244704395, "f1_macro": 0.8937572884194912, "f1_weighted": 0.8989465228549908, "precision": 0.8946194192964826, "recall": 0.8930720068510613 }
google/flan-t5-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "out_proj" ]
225,218,421
1,714,280
8
2640.98
1715.72
{ "accuracy": 0.5288491938033513, "f1_macro": 0.4391417107422545, "f1_weighted": 0.49398884315574265, "precision": 0.48366043598203423, "recall": 0.4656912252805828 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k_proj", "q_proj", "v_proj" ]
490,819,853
83,463,424
4
1682.77
1239.16
{ "accuracy": 0.9018337021814733, "f1_macro": 0.8973518270033878, "f1_weighted": 0.901926763806326, "precision": 0.8983431633289526, "recall": 0.8964917421940851 }
google/flan-t5-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k", "o", "q", "v", "wi_0", "wi_1", "wo" ]
226,914,165
3,410,024
8
2684.9
1725.93
{ "accuracy": 0.5288491938033513, "f1_macro": 0.4370755285845007, "f1_weighted": 0.4946075324705599, "precision": 0.49482949226226025, "recall": 0.4584299106614047 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k_proj", "q_proj", "v_proj" ]
491,704,589
84,348,160
8
1688.75
1255.55
{ "accuracy": 0.9037306354726525, "f1_macro": 0.8994046606341128, "f1_weighted": 0.903815252990923, "precision": 0.9003330113724781, "recall": 0.8986246033127124 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k_proj", "q_proj", "v_proj" ]
493,474,061
86,117,632
16
1700.69
1253.9
{ "accuracy": 0.901280429971546, "f1_macro": 0.8969377414857906, "f1_weighted": 0.9013812193718886, "precision": 0.8981345585782979, "recall": 0.8959601687765681 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k_proj", "q_proj", "v_proj" ]
497,013,005
89,656,576
32
1724.58
1250.48
{ "accuracy": 0.8977236800505849, "f1_macro": 0.8929478798534165, "f1_weighted": 0.8978605616617547, "precision": 0.8933585360821303, "recall": 0.892692971979935 }
google/flan-t5-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "dense", "out_proj" ]
226,932,701
3,428,560
16
2686.26
1722.61
{ "accuracy": 0.7421751501738856, "f1_macro": 0.6502244728766655, "f1_weighted": 0.7159253112324739, "precision": 0.7405507369342138, "recall": 0.6782645772149173 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k_proj", "q_proj", "v_proj" ]
504,090,893
96,734,464
64
1772.57
1257.79
{ "accuracy": 0.8964590578564654, "f1_macro": 0.8912967448549094, "f1_weighted": 0.896556411120817, "precision": 0.8910517534139267, "recall": 0.8916823210117144 }
google/flan-t5-base
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k", "o", "q", "v", "wi_0", "wi_1", "wo" ]
230,324,189
6,820,048
16
2726.74
1731.62
{ "accuracy": 0.7667562440720834, "f1_macro": 0.6781852408077077, "f1_weighted": 0.7409036527368469, "precision": 0.7470279056362396, "recall": 0.702489786974115 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k_proj", "q_proj", "v_proj" ]
518,246,669
110,890,240
128
1868.57
1263.44
{ "accuracy": 0.8917167246285173, "f1_macro": 0.8851692432395832, "f1_weighted": 0.8918644725757472, "precision": 0.8839875483112685, "recall": 0.8865963690779248 }
RUCAIBox/mvp
50,775
12,652
{ "auto_find_batch_size": true, "gradient_accumulation_steps": 4, "learning_rate": 0.00005, "logging_steps": 1, "lr_scheduler_type": "linear", "num_train_epochs": 1, "optim": "adamw_8bit", "output_dir": "outputs", "report_to": "none", "save_strategy": "no", "save_total_limit": 0, "seed": 3407, ...
[ "k_proj", "q_proj", "v_proj" ]
546,558,221
139,201,792
256
2060.56
1267.53
{ "accuracy": 0.8906892190957951, "f1_macro": 0.8846087919705622, "f1_weighted": 0.8907712876699946, "precision": 0.8834965546483843, "recall": 0.8859103010451855 }