easyword-model-peft-distilled-1.3B

This model is a fine-tuned version of facebook/nllb-200-distilled-1.3B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.1607
• Bleu: 0.0
• Gen Len: 5.9876

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 16
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
No log	1.0	31	4.9650	0.0	9.3975
No log	2.0	62	4.4353	0.1522	8.6957
No log	3.0	93	3.8967	1.2792	6.8137
No log	4.0	124	3.6053	2.6004	6.0062
No log	5.0	155	3.5239	2.9339	5.8571
No log	6.0	186	3.4692	2.6031	5.8261
No log	7.0	217	3.4244	2.6536	5.795
No log	8.0	248	3.3865	2.6445	5.8509
No log	9.0	279	3.3555	2.5482	5.9193
No log	10.0	310	3.3325	3.087	5.913
No log	11.0	341	3.3141	3.3511	5.9006
No log	12.0	372	3.2986	3.3511	5.8944
No log	13.0	403	3.2871	3.9871	5.8758
No log	14.0	434	3.2787	3.3083	5.882
No log	15.0	465	3.2738	3.3083	5.882
No log	16.0	496	3.2720	3.3083	5.882

Framework versions

• Transformers 4.29.2
• Pytorch 2.0.1+cu117
• Datasets 2.12.0
• Tokenizers 0.13.3