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/b943856b3535c48c_train_data.json
 model_type: AutoModelForCausalLM
@@ -89,7 +89,7 @@ wandb_name: 31479f65-afec-4329-9da8-252bb848872b
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 31479f65-afec-4329-9da8-252bb848872b
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -100,6 +100,8 @@ xformers_attention: null
 # 22038c2d-0745-46b6-b12d-6a3972df7b59
 
 This model is a fine-tuned version of [oopsung/llama2-7b-n-ox-test-v1](https://huggingface.co/oopsung/llama2-7b-n-ox-test-v1) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.3297
 
 ## Model description
 
@@ -126,14 +128,19 @@ 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        | 0.0002 | 1    | 1.8617          |
+| 1.6818        | 0.0017 | 10   | 1.7342          |
+| 1.3906        | 0.0035 | 20   | 1.4662          |
+| 1.3293        | 0.0052 | 30   | 1.3750          |
+| 1.2272        | 0.0069 | 40   | 1.3357          |
+| 1.3169        | 0.0087 | 50   | 1.3297          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6bb6a9bb5c752c518d571a2cf6e72f4a04c3b6eef119eb86b6747f601e93c9c1
+oid sha256:808551afc7164a12d29d6b2b5e5aaf87c673619d84a83ce8a5e13760137ac1d5
 size 80115210