End of training

README.md

@@ -65,7 +65,7 @@ lora_model_dir: null
 lora_r: 8
 lora_target_linear: true
 lr_scheduler: cosine
-max_steps: 1
+max_steps: 50
 micro_batch_size: 8
 mlflow_experiment_name: /tmp/b28efcc52b4fb0ef_train_data.json
 model_type: AutoModelForCausalLM
@@ -90,7 +90,7 @@ wandb_name: 0d3fb56b-7db3-47b5-a5f9-c48b35245476
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 0d3fb56b-7db3-47b5-a5f9-c48b35245476
-warmup_steps: 1
+warmup_steps: 2
 weight_decay: 0.0
 xformers_attention: null
 
@@ -101,6 +101,8 @@ xformers_attention: null
 # 0d3fb56b-7db3-47b5-a5f9-c48b35245476
 
 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: 0.0036
 
 ## Model description
 
@@ -128,13 +130,18 @@ The following hyperparameters were used during training:
 - 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
+- training_steps: 20
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
 | No log        | 0.0503 | 1    | 4.6239          |
+| No log        | 0.2013 | 4    | 2.7055          |
+| No log        | 0.4025 | 8    | 0.3412          |
+| 2.1898        | 0.6038 | 12   | 0.0100          |
+| 2.1898        | 0.8050 | 16   | 0.0039          |
+| 0.0158        | 1.0126 | 20   | 0.0036          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:954bcb3eed26a500b34ab63a547398491c30bf2832f951460d0c39a894f6b3cb
+oid sha256:fb5049d65155da70c01f44b043313cc326a31138820dc163e24da91dceb4ec72
 size 84047370