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/57a9c77b8cccbc26_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 2b7935c7-28db-4bcd-a075-490b9803b3cc
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 2b7935c7-28db-4bcd-a075-490b9803b3cc
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 2349d5fb-7778-e934-bfb3-25da4637e04a
 
 This model is a fine-tuned version of [NousResearch/Hermes-2-Pro-Llama-3-8B](https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.7958
 
 ## 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        | 0.0006 | 1    | 3.3459          |
+| 3.119         | 0.0061 | 10   | 3.2056          |
+| 3.0813        | 0.0122 | 20   | 2.9366          |
+| 2.5708        | 0.0183 | 30   | 2.8241          |
+| 2.8239        | 0.0244 | 40   | 2.7999          |
+| 2.8232        | 0.0305 | 50   | 2.7958          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5a2bb289a5b09f4d1bbd393d11fe00511ad95357a45a0caf6e072aec2989e030
+oid sha256:5765bedf0b95c65a682b1a6a25c1727f6a138feaad62a3eb9d75131836c84638
 size 84047370