End of training

README.md

@@ -66,7 +66,7 @@ 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/83085abdb452fabf_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 7f3f9b8d-829f-4c88-9b9e-aaae53881979
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 7f3f9b8d-829f-4c88-9b9e-aaae53881979
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 98307d26-8b1a-42bb-9db6-7c70af68ef06
 
 This model is a fine-tuned version of [unsloth/codellama-7b](https://huggingface.co/unsloth/codellama-7b) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.3620
 
 ## Model description
 
@@ -128,14 +130,19 @@ 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    | 4.1863          |
+| 4.3793        | 0.0017 | 10   | 3.8276          |
+| 2.9154        | 0.0034 | 20   | 2.8750          |
+| 2.6043        | 0.0051 | 30   | 2.4660          |
+| 2.25          | 0.0068 | 40   | 2.3746          |
+| 2.3297        | 0.0085 | 50   | 2.3620          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bb0fd7845762496eee5d0265951b012cef92c7acdfb675b252be06209f3dffc0
+oid sha256:bfde245d3e4a89617c6dda11ca4bb4a078376e1d97d020f1a1a0e18c44ff17b9
 size 80115210