Model save

README.md

@@ -55,7 +55,7 @@ wandb_log_model:
 
 gradient_accumulation_steps: 1
 micro_batch_size: 1
-num_epochs: 2
+num_epochs: 3
 optimizer: adamw_torch
 lr_scheduler: cosine
 learning_rate: 0.00002
@@ -118,7 +118,7 @@ auto_resume_from_checkpoints: true
 
 This model is a fine-tuned version of [Qwen/Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the medalpaca/medical_meadow_medqa dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.1257
+- Loss: 0.1238
 
 ## Model description
 
@@ -147,8 +147,8 @@ The following hyperparameters were used during training:
 - total_eval_batch_size: 4
 - optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: cosine
-- lr_scheduler_warmup_steps: 4
-- num_epochs: 2
+- lr_scheduler_warmup_steps: 6
+- num_epochs: 3
 
 ### Training results
 
@@ -162,6 +162,10 @@ The following hyperparameters were used during training:
 | 0.1228        | 1.5   | 108  | 0.1263          |
 | 0.1199        | 1.75  | 126  | 0.1260          |
 | 0.1393        | 2.0   | 144  | 0.1257          |
+| 0.1146        | 2.25  | 162  | 0.1244          |
+| 0.1161        | 2.5   | 180  | 0.1238          |
+| 0.139         | 2.75  | 198  | 0.1238          |
+| 0.0927        | 3.0   | 216  | 0.1238          |
 
 
 ### Framework versions