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.

moe_tur_multi_lr_x2

This model is a fine-tuned version of on the arrow 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: 0.0002
train_batch_size: 8
eval_batch_size: 32
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 9561
training_steps: 95611
mixed_precision_training: Native AMP

Training Loss	Epoch	Step	Validation Loss
No log	0	0	10.9641
4.6362	1.0458	10000	4.6958
3.8819	2.0916	20000	4.1232
3.4131	3.1374	30000	3.9853
2.9345	4.1832	40000	4.0382
2.4231	5.2290	50000	4.2603
1.8596	6.2749	60000	4.6337
1.3436	7.3207	70000	5.1018
0.9107	8.3665	80000	5.5970
0.5969	9.4123	90000	5.9656
0.5919	9.9991	95611	6.0199

Safetensors

Model size

0.5B params

Tensor type

F32