End of training

README.md

@@ -15,8 +15,6 @@ should probably proofread and complete it, then remove this comment. -->
 # bert_pretrained
 
 This model is a fine-tuned version of [temporary0-0name/bert_pretrained](https://huggingface.co/temporary0-0name/bert_pretrained) on the wikitext dataset.
-It achieves the following results on the evaluation set:
-- Loss: 7.6706
 
 ## Model description
 
@@ -39,20 +37,16 @@ The following hyperparameters were used during training:
 - train_batch_size: 32
 - eval_batch_size: 32
 - seed: 42
-- gradient_accumulation_steps: 2
-- total_train_batch_size: 64
+- gradient_accumulation_steps: 8
+- total_train_batch_size: 256
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: cosine
-- lr_scheduler_warmup_steps: 10
+- lr_scheduler_warmup_steps: 100
 - num_epochs: 1
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss |
-|:-------------:|:-----:|:----:|:---------------:|
-| 8.2487        | 0.5   | 10   | 7.8067          |
-| 7.174         | 1.0   | 20   | 7.6706          |
 
 
 ### Framework versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9946fef6200fa07df426b0abeb44ef8f28676f48907c515638f6fc4065a84e5c
+oid sha256:bd7bf0ed21c606c982b8a8a1fd570f24dbcfbdefb1de049ce9cc00f774dd115d
 size 438080896

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fab70017f5e66b35832ddcea17a307f9b23d5c51cbc3e7e3e5881110c18c025b
+oid sha256:3d0e09faf58532b4fd9e45759e31e5d358eecbd919d15e7901d3071f58b7de7b
 size 4091