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/e51bf03bb498f9d1_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 321707c0-4ea8-4130-bb14-e36f4b768cbb
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 321707c0-4ea8-4130-bb14-e36f4b768cbb
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 321707c0-4ea8-4130-bb14-e36f4b768cbb
 
 This model is a fine-tuned version of [NousResearch/Hermes-3-Llama-3.1-8B](https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.1833
 
 ## Model description
 
@@ -128,14 +130,18 @@ 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: 13
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss |
 |:-------------:|:-----:|:----:|:---------------:|
 | No log        | 0.08  | 1    | 5.9565          |
+| No log        | 0.24  | 3    | 5.9437          |
+| 6.0911        | 0.48  | 6    | 5.2834          |
+| 6.0911        | 0.72  | 9    | 3.5568          |
+| 4.2865        | 0.96  | 12   | 2.1833          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ddc6fd6342d4ae35d41b9ec288c5cb044a804552afea84d0005ccadc632375f
+oid sha256:4de70bb5a77fa1beae11dd9b272792c80b5fdc6f3525db74c513911b6bc9af0f
 size 84047370