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/a2f34f1ba0195eb3_train_data.json
 model_type: AutoModelForCausalLM
@@ -92,7 +92,7 @@ wandb_name: 66d60cd9-b42c-4bd6-8fec-a13c2c330f27
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 66d60cd9-b42c-4bd6-8fec-a13c2c330f27
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -103,6 +103,8 @@ xformers_attention: null
 # 66d60cd9-b42c-4bd6-8fec-a13c2c330f27
 
 This model is a fine-tuned version of [NousResearch/CodeLlama-13b-hf](https://huggingface.co/NousResearch/CodeLlama-13b-hf) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 4.8656
 
 ## Model description
 
@@ -129,14 +131,18 @@ 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: 12
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
 | No log        | 0.0870 | 1    | 8.6576          |
+| No log        | 0.2609 | 3    | 8.6390          |
+| 37.8497       | 0.5217 | 6    | 8.4498          |
+| 37.8497       | 0.7826 | 9    | 7.2075          |
+| 36.0964       | 1.0435 | 12   | 4.8656          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:01febafffb8abe00f9cea7fe56b5d1caf7c5da48f3d44e2351edc6b62481a3e4
+oid sha256:8b0b8b409cd32b19d5a89fb11da03256c2db834349988883e44bdf1186eca3cf
 size 125375434