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/3229beaaf30f09a3_train_data.json
 model_type: AutoModelForCausalLM
@@ -92,7 +92,7 @@ wandb_name: 6f876187-6d0a-4410-b9c7-16623cec1fad
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 6f876187-6d0a-4410-b9c7-16623cec1fad
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -103,6 +103,8 @@ xformers_attention: null
 # ea7b14e5-4610-46ac-b82c-e7e7b273e0ed
 
 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: 0.6734
 
 ## Model description
 
@@ -129,14 +131,19 @@ 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        | 0.0037 | 1    | 6.8450          |
+| 4.2657        | 0.0366 | 10   | 1.4937          |
+| 0.6925        | 0.0732 | 20   | 0.7364          |
+| 0.6899        | 0.1098 | 30   | 0.6869          |
+| 0.631         | 0.1464 | 40   | 0.6842          |
+| 0.7376        | 0.1830 | 50   | 0.6734          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:35efc1adc6fb50377cd67d69cd7a3d6b1f3ab0f12fe8ded5ab34993b77d0dc17
+oid sha256:a766e8dabe4311da594525419753058040d2b764e5c8837b9072850a2782d932
 size 80115210