End of training

README.md

@@ -66,7 +66,7 @@ lora_model_dir: null
 lora_r: 8
 lora_target_linear: true
 lr_scheduler: cosine
-max_steps: 1
+max_steps: 50
 micro_batch_size: 2
 mlflow_experiment_name: /tmp/8e867e0c95f85985_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 3d129001-be65-406a-a409-10f99a17a16f
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 3d129001-be65-406a-a409-10f99a17a16f
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 3d129001-be65-406a-a409-10f99a17a16f
 
 This model is a fine-tuned version of [Qwen/Qwen2.5-1.5B](https://huggingface.co/Qwen/Qwen2.5-1.5B) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.0142
 
 ## Model description
 
@@ -128,14 +130,19 @@ The following hyperparameters were used during training:
 - total_train_batch_size: 8
 - optimizer: Use OptimizerNames.ADAMW_BNB 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: 2
-- training_steps: 1
+- lr_scheduler_warmup_steps: 10
+- training_steps: 50
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
 | No log        | 0.0004 | 1    | 7.8545          |
+| 7.296         | 0.0043 | 10   | 6.5557          |
+| 3.2915        | 0.0085 | 20   | 3.1997          |
+| 3.42          | 0.0128 | 30   | 2.4580          |
+| 2.6229        | 0.0170 | 40   | 2.0688          |
+| 1.1089        | 0.0213 | 50   | 2.0142          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:83b54464d5df3f1d5f7a4e57dd0b7a166ec46ed44f611b61f9b2d2212d264950
+oid sha256:33ddf841d4b71b7d4b1f22145c8d510d0acddafb137e85633ac728f7f932c88e
 size 37070634