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/e00944ee01ce6880_train_data.json
 model_type: AutoModelForCausalLM
@@ -93,7 +93,7 @@ wandb_name: 040e49b4-463c-4ada-a9fe-3e7135c1ee8d
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 040e49b4-463c-4ada-a9fe-3e7135c1ee8d
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -104,6 +104,8 @@ xformers_attention: null
 # e3623524-fb69-26d8-9a19-f2016b738f37
 
 This model is a fine-tuned version of [NousResearch/Llama-3.2-1B](https://huggingface.co/NousResearch/Llama-3.2-1B) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0601
 
 ## Model description
 
@@ -130,14 +132,19 @@ 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        | 0.0000 | 1    | 0.7639          |
+| 0.6406        | 0.0004 | 10   | 0.4751          |
+| 0.1668        | 0.0008 | 20   | 0.1123          |
+| 0.0929        | 0.0012 | 30   | 0.0808          |
+| 0.0734        | 0.0016 | 40   | 0.0629          |
+| 0.0592        | 0.0020 | 50   | 0.0601          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a16a69d2f9ef8c4f79b5532337dc26527dbfc241e0c1058f515b348210a5ee7f
+oid sha256:73d1873c653499f10064b86b301894ce9a6cce762eeb6b53ecf590f07c339b18
 size 22624970