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: 2
 mlflow_experiment_name: /tmp/15160cecae42f4a0_train_data.json
 model_type: AutoModelForCausalLM
@@ -90,7 +90,7 @@ wandb_name: 881d25a8-b8b9-4102-ad7e-27d5b99cca9f
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 881d25a8-b8b9-4102-ad7e-27d5b99cca9f
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -101,6 +101,8 @@ xformers_attention: null
 # 6bbfbc35-f0d5-427d-aac1-55d07a905c7d
 
 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: 0.7292
 
 ## Model description
 
@@ -127,14 +129,19 @@ 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    | 1.8974          |
+| 1.5286        | 0.0007 | 10   | 1.6220          |
+| 0.924         | 0.0015 | 20   | 0.8477          |
+| 0.919         | 0.0022 | 30   | 0.7541          |
+| 0.673         | 0.0030 | 40   | 0.7331          |
+| 0.8729        | 0.0037 | 50   | 0.7292          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:db449639f8dcdc61bdad33c1d1032ec66a622bd90f49e69e9857945a49537e7e
+oid sha256:1f199dc1b4274d5d20302d6c7479e31e45e6e958e1fc5b6527233e213a31b0b4
 size 84047370