learning_rate: 0.0001
train_batch_size: 32
eval_batch_size: 64
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 30
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu 1 Msl	Bleu 2 Msl	Bleu 3 Msl	Bleu 4 Msl	Ter Msl	Ter Asl
No log	1.0	75	0.9146	0.5633	0.0137	0.0042	0.0021	{'score': 244.91129785247435, 'num_edits': 2623, 'ref_length': 1071.0}	100
No log	2.0	150	0.7602	0.58	0.0139	0.0043	0.0022	{'score': 63.86554621848739, 'num_edits': 684, 'ref_length': 1071.0}	100
No log	3.0	225	0.7428	0.6433	0.0147	0.0044	0.0022	{'score': 16.900093370681606, 'num_edits': 181, 'ref_length': 1071.0}	100
No log	4.0	300	0.6946	0.6567	0.0148	0.0044	0.0022	{'score': 16.713352007469652, 'num_edits': 179, 'ref_length': 1071.0}	100
No log	5.0	375	0.6935	0.6133	0.0143	0.0043	0.0022	{'score': 18.674136321195146, 'num_edits': 200, 'ref_length': 1071.0}	100
No log	6.0	450	0.7201	0.63	0.0145	0.0044	0.0022	{'score': 17.273576097105508, 'num_edits': 185, 'ref_length': 1071.0}	100
0.6196	7.0	525	0.7572	0.6333	0.0146	0.0044	0.0022	{'score': 18.860877684407097, 'num_edits': 202, 'ref_length': 1071.0}	100
0.6196	8.0	600	0.7963	0.62	0.0144	0.0043	0.0022	{'score': 17.553688141923434, 'num_edits': 188, 'ref_length': 1071.0}	100
0.6196	9.0	675	0.7047	0.6333	0.0146	0.0044	0.0022	{'score': 17.366946778711483, 'num_edits': 186, 'ref_length': 1071.0}	100
0.6196	10.0	750	0.6648	0.62	0.0144	0.0043	0.0022	{'score': 17.92717086834734, 'num_edits': 192, 'ref_length': 1071.0}	100
0.6196	11.0	825	0.8197	0.62	0.0144	0.0043	0.0022	{'score': 18.11391223155929, 'num_edits': 194, 'ref_length': 1071.0}	100
0.6196	12.0	900	0.8325	0.6	0.0142	0.0043	0.0022	{'score': 18.020541549953315, 'num_edits': 193, 'ref_length': 1071.0}	100
0.6196	13.0	975	0.7735	0.5633	0.0137	0.0042	0.0021	{'score': 17.92717086834734, 'num_edits': 192, 'ref_length': 1071.0}	100
0.0502	14.0	1050	0.7875	0.59	0.0140	0.0043	0.0022	{'score': 16.433239962651726, 'num_edits': 176, 'ref_length': 1071.0}	100
0.0502	15.0	1125	0.7229	0.57	0.0138	0.0042	0.0022	{'score': 16.713352007469652, 'num_edits': 179, 'ref_length': 1071.0}	100
0.0502	16.0	1200	0.7772	0.65	0.0147	0.0044	0.0022	{'score': 14.84593837535014, 'num_edits': 159, 'ref_length': 1071.0}	100
0.0502	17.0	1275	0.7100	0.6267	0.0145	0.0044	0.0022	{'score': 15.966386554621847, 'num_edits': 171, 'ref_length': 1071.0}	100
0.0502	18.0	1350	0.7534	0.6333	0.0146	0.0044	0.0022	{'score': 14.752567693744165, 'num_edits': 158, 'ref_length': 1071.0}	100
0.0502	19.0	1425	0.7945	0.6067	0.0142	0.0043	0.0022	{'score': 16.5266106442577, 'num_edits': 177, 'ref_length': 1071.0}	100
0.022	20.0	1500	0.7580	0.63	0.0145	0.0044	0.0022	{'score': 15.966386554621847, 'num_edits': 171, 'ref_length': 1071.0}	100
0.022	21.0	1575	0.7658	0.5933	0.0141	0.0043	0.0022	{'score': 16.80672268907563, 'num_edits': 180, 'ref_length': 1071.0}	100
0.022	22.0	1650	0.7830	0.6167	0.0144	0.0043	0.0022	{'score': 16.80672268907563, 'num_edits': 180, 'ref_length': 1071.0}	100
0.022	23.0	1725	0.7999	0.63	0.0145	0.0044	0.0022	{'score': 15.779645191409896, 'num_edits': 169, 'ref_length': 1071.0}	100
0.022	24.0	1800	0.8010	0.5967	0.0141	0.0043	0.0022	{'score': 16.900093370681606, 'num_edits': 181, 'ref_length': 1071.0}	100
0.022	25.0	1875	0.8060	0.6267	0.0145	0.0044	0.0022	{'score': 16.5266106442577, 'num_edits': 177, 'ref_length': 1071.0}	100
0.022	26.0	1950	0.7953	0.6167	0.0144	0.0043	0.0022	{'score': 16.33986928104575, 'num_edits': 175, 'ref_length': 1071.0}	100
0.0146	27.0	2025	0.7977	0.6133	0.0143	0.0043	0.0022	{'score': 16.246498599439775, 'num_edits': 174, 'ref_length': 1071.0}	100
0.0146	28.0	2100	0.8028	0.6133	0.0143	0.0043	0.0022	{'score': 16.33986928104575, 'num_edits': 175, 'ref_length': 1071.0}	100
0.0146	29.0	2175	0.8061	0.6133	0.0143	0.0043	0.0022	{'score': 16.5266106442577, 'num_edits': 177, 'ref_length': 1071.0}	100
0.0146	30.0	2250	0.8030	0.6133	0.0143	0.0043	0.0022	{'score': 16.246498599439775, 'num_edits': 174, 'ref_length': 1071.0}	100

Framework versions

Transformers 4.49.0
Pytorch 2.5.1+cu124
Datasets 3.3.2
Tokenizers 0.21.0

vania2911/exp2_10partition_modeloorig

exp2_10partition_modeloorig

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

vania2911/
exp2_10partition_modeloorig