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/6bc7254c9c745e8a_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 17b1db57-7be4-4280-bf23-2cbc8e36894b
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 17b1db57-7be4-4280-bf23-2cbc8e36894b
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 1bdb6a06-bc6d-4a86-b0e6-f706b26cfc31
 
 This model is a fine-tuned version of [WhiteRabbitNeo/Llama-3-WhiteRabbitNeo-8B-v2.0](https://huggingface.co/WhiteRabbitNeo/Llama-3-WhiteRabbitNeo-8B-v2.0) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.3181
 
 ## 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        | 0.0000 | 1    | 0.4565          |
+| 0.4417        | 0.0004 | 10   | 0.3890          |
+| 0.3235        | 0.0007 | 20   | 0.3387          |
+| 0.329         | 0.0011 | 30   | 0.3223          |
+| 0.3237        | 0.0015 | 40   | 0.3194          |
+| 0.2629        | 0.0018 | 50   | 0.3181          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6d0b524bb210794e79ba458925dfcae18470cdce2b7c40a7b424774a436eb70e
+oid sha256:9c00a58ca7ecfbc652435f9df402af3c1a25f4edec5f8a54a79b89436fc94691
 size 84047370