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/7cf24425fc8c1fe3_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: fb87b24e-3181-48bb-aa49-a1bef5ce55ba
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: fb87b24e-3181-48bb-aa49-a1bef5ce55ba
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # a24cac83-40ac-428a-b8e3-d6a767c9c5b1
 
 This model is a fine-tuned version of [unsloth/codegemma-2b](https://huggingface.co/unsloth/codegemma-2b) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.1534
 
 ## 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        | 0.0001 | 1    | 0.2945          |
+| 0.2814        | 0.0007 | 10   | 0.2350          |
+| 0.2037        | 0.0015 | 20   | 0.1696          |
+| 0.1769        | 0.0022 | 30   | 0.1588          |
+| 0.1775        | 0.0030 | 40   | 0.1540          |
+| 0.1446        | 0.0037 | 50   | 0.1534          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f0131990b7a29cc442412c63c72cc2ba0fbc311f29626a35835bb1ed4db59ed0
+oid sha256:c914b70683579e5eca8ca6246067d602ace2f5f47f28cd1dc5dfa31a20434b22
 size 39313978