End of training

README.md

@@ -76,7 +76,7 @@ Trained on 145,744,973 tokens from the [wikimedia/wikipedia](https://huggingface
 # Training Objective
 
 ```
-DistillationObjective(logits_loss_component=LossComponent(label=logits, weight=1, loss_fn=kl), attn_loss_component=LossComponent(label=attn, weight=25.0, loss_fn=raw_mse, layer_mapper=layer-2, projector=linear))
+DistillationObjective(logits_loss_component=LossComponent(label=logits, weight=1, loss_fn=kl), attn_loss_component=LossComponent(label=attn, weight=25.0, loss_fn=cos, layer_mapper=last, projector=linear))
 ```
 
 # Hyperparameters
@@ -93,9 +93,9 @@ The following hyperparameters were used during training:
 - lr_scheduler_type: `cosine_with_min_lr`
 - lr_scheduler_warmup_ratio: `0.5`
 - num_epochs: `1.0`
-- distillation_objective: `DistillationObjective(logits_loss_component=LossComponent(label=logits, weight=1, loss_fn=kl), attn_loss_component=LossComponent(label=attn, weight=25.0, loss_fn=raw_mse, layer_mapper=layer-2, projector=linear))`
+- distillation_objective: `DistillationObjective(logits_loss_component=LossComponent(label=logits, weight=1, loss_fn=kl), attn_loss_component=LossComponent(label=attn, weight=25.0, loss_fn=cos, layer_mapper=last, projector=linear))`
 - train_embeddings: `True`
-- lr_scheduler: `<torch.optim.lr_scheduler.LambdaLR object at 0x7fa4ed4dfeb0>`
+- lr_scheduler: `<torch.optim.lr_scheduler.LambdaLR object at 0x7fa4b4615600>`
 - student_model_name_or_path: `None`
 - student_config_name_or_path: `None`
 - student_model_config: `None`

logs/attn_layer_mapper=last, attn_loss_fn=cos, attn_projector=linear/events.out.tfevents.1724580358.e3f806ea38c9

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:437e888f2d276229ad4026dd0f8567166b02a24ff896ba3d9a254b658681628e
+size 29625282

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:42191dab9a4b9687aabcaafa9da63cdec7ff009efc613b79337390f83c04f671
+oid sha256:2ed43a88fc7e349c7bf9ba8621d0003dc6392f06e7b9617f5e43ca439de48048
 size 248894656

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9803d0ad587257ccf7e239fb6144fa5011ca6c9c30a32e245bcde8c65cce0d65
+oid sha256:7978e7e7ca9306a3920b3495776044f0c00090e252b573798e8c9b3d514e8489
 size 5432