1b4e9f184a48d3310a41de2a87465281

This model is a fine-tuned version of google-bert/bert-base-multilingual-uncased on the nyu-mll/glue [sst2] dataset. It achieves the following results on the evaluation set:

• Loss: 0.4241
• Data Size: 1.0
• Epoch Runtime: 105.4318
• Accuracy: 0.8796
• F1 Macro: 0.8793
• Rouge1: 0.8808
• Rouge2: 0.0
• Rougel: 0.8796
• Rougelsum: 0.8796

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• total_train_batch_size: 32
• total_eval_batch_size: 32
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss	Data Size	Epoch Runtime	Accuracy	F1 Macro	Rouge1	Rouge2	Rougel	Rougelsum
No log	0	0	0.6953	0	0.9821	0.5093	0.3374	0.5093	0.0	0.5093	0.5081
No log	1	2104	0.6842	0.0078	2.6533	0.5093	0.3374	0.5093	0.0	0.5093	0.5081
No log	2	4208	0.5368	0.0156	2.7749	0.7650	0.7614	0.7650	0.0	0.7662	0.7650
0.012	3	6312	0.5200	0.0312	4.7773	0.7546	0.7447	0.7546	0.0	0.7558	0.7535
0.451	4	8416	0.3855	0.0625	7.8871	0.8229	0.8226	0.8241	0.0	0.8229	0.8218
0.344	5	10520	0.3623	0.125	14.1505	0.8438	0.8433	0.8438	0.0	0.8438	0.8438
0.2507	6	12624	0.3806	0.25	28.0165	0.8542	0.8539	0.8542	0.0	0.8553	0.8542
0.2196	7	14728	0.3578	0.5	53.4898	0.8704	0.8696	0.8704	0.0	0.8704	0.8704
0.176	8.0	16832	0.4344	1.0	106.4276	0.8426	0.8418	0.8426	0.0	0.8426	0.8426
0.1415	9.0	18936	0.4417	1.0	105.4077	0.8681	0.8678	0.8681	0.0	0.8681	0.8681
0.1423	10.0	21040	0.3722	1.0	106.3322	0.8634	0.8631	0.8634	0.0	0.8634	0.8634
0.1177	11.0	23144	0.4241	1.0	105.4318	0.8796	0.8793	0.8808	0.0	0.8796	0.8796

Framework versions

• Transformers 4.57.0
• Pytorch 2.8.0+cu128
• Datasets 4.3.0
• Tokenizers 0.22.1