End of training

README.md

@@ -64,7 +64,7 @@ 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/c57750dcb7647266_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: ff959fe8-05d4-40a9-9620-d274aabd4629
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: ff959fe8-05d4-40a9-9620-d274aabd4629
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # ff959fe8-05d4-40a9-9620-d274aabd4629
 
 This model is a fine-tuned version of [Korabbit/llama-2-ko-7b](https://huggingface.co/Korabbit/llama-2-ko-7b) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.2011
 
 ## 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        | 0.0009 | 1    | 2.8427          |
+| 2.427         | 0.0088 | 10   | 1.8639          |
+| 1.376         | 0.0176 | 20   | 1.3033          |
+| 1.3645        | 0.0264 | 30   | 1.2372          |
+| 1.2459        | 0.0352 | 40   | 1.2057          |
+| 1.2934        | 0.0440 | 50   | 1.2011          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:67f5ecbc14c680aeaf4d45d4d93cc0963afeb7b7acc281309e5dbb4443905560
+oid sha256:fd3e969196229d12d08f295b6c7dd302eae13e9ae0e03bfc24512ad89f04c3db
 size 80115210