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/f07dc76a41dc794d_train_data.json
 model_type: AutoModelForCausalLM
@@ -90,7 +90,7 @@ wandb_name: b326d613-bc5b-4ca9-8960-9dffd61c55dd
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: b326d613-bc5b-4ca9-8960-9dffd61c55dd
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -101,6 +101,8 @@ xformers_attention: null
 # b326d613-bc5b-4ca9-8960-9dffd61c55dd
 
 This model is a fine-tuned version of [The-matt/llama2_ko-7b_distinctive-snowflake-182_1060](https://huggingface.co/The-matt/llama2_ko-7b_distinctive-snowflake-182_1060) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 3.3709
 
 ## 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        | 0.0017 | 1    | 4.2003          |
+| 3.9812        | 0.0175 | 10   | 3.9207          |
+| 3.5651        | 0.0350 | 20   | 3.5672          |
+| 3.3628        | 0.0524 | 30   | 3.4137          |
+| 3.5705        | 0.0699 | 40   | 3.3751          |
+| 3.2939        | 0.0874 | 50   | 3.3709          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6319c6348e3c9bb8f553724c2bb7d5f03fdbcc92ded20bd057540f63cd5b2439
+oid sha256:e748071d1db838f6bcd79532786d713c36c611b585198b78f7338454c8870a8e
 size 80115210