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/35da456c8da4e9c2_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: cf342ba4-ce68-47ec-923b-899f2f1d9a3d
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: cf342ba4-ce68-47ec-923b-899f2f1d9a3d
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # b9647a5b-eae5-4dad-8785-a8cfdfcc0236
 
 This model is a fine-tuned version of [unsloth/codegemma-7b-it](https://huggingface.co/unsloth/codegemma-7b-it) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.4959
 
 ## 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        | 0.0018 | 1    | 2.8794          |
+| 2.6699        | 0.0182 | 10   | 2.2895          |
+| 1.8973        | 0.0364 | 20   | 1.7337          |
+| 1.5553        | 0.0547 | 30   | 1.5689          |
+| 1.6779        | 0.0729 | 40   | 1.5071          |
+| 1.5104        | 0.0911 | 50   | 1.4959          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2733ef25ebc6b45298c6eee5f5bdcf336e3792156d509683aff0303e767d6672
+oid sha256:b82c54e44d14d448ab4e37c31c60293340a11ba151c60516055150dc94b265af
 size 100149034