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: 50
+max_steps: 1
 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: 10
+warmup_steps: 1
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,8 +102,6 @@ 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: 1.1999
 
 ## Model description
 
@@ -130,19 +128,14 @@ 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: 10
-- training_steps: 50
+- lr_scheduler_warmup_steps: 2
+- training_steps: 1
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
 | No log        | 0.0009 | 1    | 2.8427          |
-| 2.4219        | 0.0088 | 10   | 1.8618          |
-| 1.3786        | 0.0176 | 20   | 1.3012          |
-| 1.3621        | 0.0264 | 30   | 1.2346          |
-| 1.2442        | 0.0352 | 40   | 1.2046          |
-| 1.294         | 0.0440 | 50   | 1.1999          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2aa1f6a616e6816242b48a7ce83c6a1454755c17cffe9419770a06c841c596d4
+oid sha256:91a0aa7191dc4463551b0402bf067199463b61260ff8f625d80070093c220b8c
 size 80115210