mamba-text-classification

This model is a fine-tuned version of state-spaces/mamba-130m-hf on imdb dataset. It achieves the following results on the evaluation set:

• Loss: 0.3637
• Accuracy: 0.9454
• F1: 0.9454
• Recall: 0.9461
• Precision: 0.9447

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 8
• eval_batch_size: 32
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Recall	Precision
0.1731	0.9997	781	0.1553	0.9425	0.9427	0.9462	0.9393
0.1316	1.9994	1562	0.1970	0.9319	0.9294	0.8974	0.9639
0.0224	2.9990	2343	0.3137	0.9454	0.9455	0.9479	0.9432
0.0002	4.0	3125	0.3501	0.9449	0.9450	0.9470	0.9431
0.0004	4.9984	3905	0.3637	0.9454	0.9454	0.9461	0.9447

Framework versions

• Transformers 4.41.0
• Pytorch 2.2.2+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1