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/f1082c2b46fe7bf4_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 963b60c6-73f0-4529-bcf2-edbd4dd823fd
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 963b60c6-73f0-4529-bcf2-edbd4dd823fd
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # d2d1a7f5-8e58-cbc3-bf50-786f54c3659e
 
 This model is a fine-tuned version of [unsloth/gemma-7b-it](https://huggingface.co/unsloth/gemma-7b-it) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.0945
 
 ## 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    | 3.1598          |
+| 3.007         | 0.0017 | 10   | 2.5921          |
+| 2.4081        | 0.0035 | 20   | 2.2510          |
+| 2.099         | 0.0052 | 30   | 2.1576          |
+| 2.0199        | 0.0069 | 40   | 2.1024          |
+| 2.036         | 0.0087 | 50   | 2.0945          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d0563417c55393c0166ed91f5adc7e069ea26ad15a58d2125b49f469f71cdaba
+oid sha256:9fc0146c4f925d1c732c9973533b0161c10c65cb2ae0523e96a42186d565efd5
 size 100149034