You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

nllb-200-3.3B-fr2mos-peft

This model is a fine-tuned version of facebook/nllb-200-3.3B on an unknown dataset. It achieves the following results on the evaluation set:

More information needed

More information needed

More information needed

The following hyperparameters were used during training:

learning_rate: 1e-05
train_batch_size: 32
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 128
optimizer: Use adamw_bnb_8bit with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_ratio: 0.03
num_epochs: 3

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
1.6928	0.4196	1000	1.5518	26.8647	26.7514
1.5401	0.8391	2000	1.4211	27.9357	28.2148
1.3751	1.2584	3000	1.3408	31.8779	26.1628
1.3299	1.6780	4000	1.2908	33.2896	26.7148
1.2386	2.0973	5000	1.2616	34.239	26.1995
1.239	2.5169	6000	1.2486	34.722	26.0361
1.2374	2.9364	7000	1.2459	34.7883	25.8891

Safetensors

Model size

3B params

Tensor type

F32

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Base model

Finetuned

(30)

this model