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/4ef0a0f8f3669573_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: e3afeb2d-6ea4-4656-b0c6-05889a3e1f72
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: e3afeb2d-6ea4-4656-b0c6-05889a3e1f72
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 4ffb6bf9-b1b6-4e27-b7b5-239f0f1e612c
 
 This model is a fine-tuned version of [unsloth/Llama-3.2-1B](https://huggingface.co/unsloth/Llama-3.2-1B) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.5843
 
 ## 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        | 0.0026 | 1    | 3.0789          |
+| 3.0951        | 0.0262 | 10   | 2.9613          |
+| 2.6461        | 0.0525 | 20   | 2.7053          |
+| 2.5546        | 0.0787 | 30   | 2.6098          |
+| 2.5121        | 0.1049 | 40   | 2.5886          |
+| 2.5034        | 0.1311 | 50   | 2.5843          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:21c61fcf91d3f3818f90d2d7e1df10da749836b705b91e23a34e3ee709e4c3d7
+oid sha256:6a393290e5a78d1c70870582c19cea60f8e559cb130a966fc153a01d5bb0db5e
 size 22624970