End of training

README.md

@@ -64,7 +64,7 @@ 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/98e8d3e5dc84aa2b_train_data.json
 model_type: AutoModelForCausalLM
@@ -89,7 +89,7 @@ wandb_name: fd2b1ff6-571b-42b0-9977-042469a19fb8
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: fd2b1ff6-571b-42b0-9977-042469a19fb8
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -100,6 +100,8 @@ xformers_attention: null
 # 5a17e82d-6f5f-9fd6-a584-cc2d17416988
 
 This model is a fine-tuned version of [unsloth/Hermes-3-Llama-3.1-8B](https://huggingface.co/unsloth/Hermes-3-Llama-3.1-8B) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.4339
 
 ## Model description
 
@@ -126,14 +128,19 @@ 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        | 0.0005 | 1    | 3.7716          |
+| 3.7568        | 0.0047 | 10   | 3.2308          |
+| 1.8883        | 0.0095 | 20   | 1.7703          |
+| 1.545         | 0.0142 | 30   | 1.4945          |
+| 1.2891        | 0.0189 | 40   | 1.4415          |
+| 1.4042        | 0.0237 | 50   | 1.4339          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9c80074b6cd6a104db8a616be2e7dd78ced8a74a7131c13546c145d42b5aaf70
+oid sha256:3dc685b53f5a59520964c1ea54507a5a0302d81ff1d7a015412ed6dc74d77368
 size 84047370