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/7246ec57ff60e750_train_data.json
 model_type: AutoModelForCausalLM
@@ -89,7 +89,7 @@ wandb_name: 3e2e85d2-5c59-408f-ad8f-6a668dd73629
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 3e2e85d2-5c59-408f-ad8f-6a668dd73629
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -100,6 +100,8 @@ xformers_attention: null
 # 304c0971-3ff3-4d22-9c80-923aa4aa7224
 
 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: 0.7871
 
 ## 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        | 0.0010 | 1    | 1.9743          |
+| 1.7926        | 0.0098 | 10   | 1.7347          |
+| 1.196         | 0.0196 | 20   | 1.0781          |
+| 0.9216        | 0.0294 | 30   | 0.8485          |
+| 0.7803        | 0.0392 | 40   | 0.7932          |
+| 0.7945        | 0.0490 | 50   | 0.7871          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c9aeafc1cda761297b54efc728a953ad1621c15c81e630c3c16d3a53d595ad81
+oid sha256:84f6e3ea9e6af6802d90bf203e3ee3858af935ce8a2e4ff7cd3b1da088fee51a
 size 80115210