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/6e6766e47fe66b56_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: e047fdab-8898-4a3b-a66d-b4ada8b849c2
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: e047fdab-8898-4a3b-a66d-b4ada8b849c2
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 5a725f0a-b269-c961-b923-dd0ee6028bef
 
 This model is a fine-tuned version of [unsloth/gemma-1.1-2b-it](https://huggingface.co/unsloth/gemma-1.1-2b-it) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.4704
 
 ## 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        | 0.0003 | 1    | 1.9688          |
+| 1.374         | 0.0035 | 10   | 1.1795          |
+| 0.884         | 0.0069 | 20   | 0.6531          |
+| 0.4755        | 0.0104 | 30   | 0.5025          |
+| 0.4492        | 0.0138 | 40   | 0.4764          |
+| 0.5057        | 0.0173 | 50   | 0.4704          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1f6c87d9405221fff4c1179fae0000200e12dcbddc5e58a0b08d836d1b3645b5
+oid sha256:c2f6284b91274da240e415c31a48ed476715bd6c29f9e5df35a9db7d453fdc3f
 size 39313978