README.md

metadata

tags:
  - generated_from_trainer
metrics:
  - accuracy
model-index:
  - name: HBERTv1_emb_compress_48_L10_H64_A2
    results: []

HBERTv1_emb_compress_48_L10_H64_A2

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 6.4103
• Accuracy: 0.1281

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: 1e-05
• train_batch_size: 96
• eval_batch_size: 96
• seed: 10
• distributed_type: multi-GPU
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 10000
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
8.6666	0.16	10000	8.5962	0.0465
7.2486	0.33	20000	7.2447	0.0467
7.0176	0.49	30000	7.0089	0.0670
6.8859	0.66	40000	6.8795	0.0840
6.7911	0.82	50000	6.7857	0.0918
6.7203	0.98	60000	6.7210	0.0952
6.6722	1.15	70000	6.6715	0.1004
6.6362	1.31	80000	6.6338	0.1033
6.5995	1.47	90000	6.6017	0.1065
6.5756	1.64	100000	6.5755	0.1092
6.5546	1.8	110000	6.5521	0.1118
6.5302	1.97	120000	6.5330	0.1138
6.5121	2.13	130000	6.5166	0.1156
6.5049	2.29	140000	6.5005	0.1174
6.483	2.46	150000	6.4869	0.1190
6.4757	2.62	160000	6.4755	0.1205
6.4659	2.79	170000	6.4655	0.1219
6.4527	2.95	180000	6.4569	0.1227
6.4517	3.11	190000	6.4477	0.1237
6.441	3.28	200000	6.4422	0.1245
6.4385	3.44	210000	6.4353	0.1253
6.4308	3.6	220000	6.4295	0.1256
6.4188	3.77	230000	6.4250	0.1264
6.422	3.93	240000	6.4213	0.1269
6.416	4.1	250000	6.4180	0.1273
6.4215	4.26	260000	6.4151	0.1276
6.4135	4.42	270000	6.4142	0.1277
6.4138	4.59	280000	6.4118	0.1281
6.41	4.75	290000	6.4097	0.1283
6.4114	4.92	300000	6.4103	0.1281

Framework versions

• Transformers 4.33.2
• Pytorch 1.14.0a0+410ce96
• Datasets 2.14.5
• Tokenizers 0.13.3