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/f44a24a0bc5385f2_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 0f3e25f3-9370-47c0-bb1f-167b56131fd8
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 0f3e25f3-9370-47c0-bb1f-167b56131fd8
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 0f3e25f3-9370-47c0-bb1f-167b56131fd8
 
 This model is a fine-tuned version of [Qwen/Qwen2.5-Math-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Math-7B-Instruct) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.2215
 
 ## 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        | 0.0008 | 1    | 2.4701          |
+| 2.4084        | 0.0081 | 10   | 2.3091          |
+| 1.7638        | 0.0163 | 20   | 1.7882          |
+| 1.4841        | 0.0244 | 30   | 1.4007          |
+| 1.2216        | 0.0326 | 40   | 1.2438          |
+| 1.1845        | 0.0407 | 50   | 1.2215          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b3e182be4fb6ff8018b15202686d5eedab72c3a1a84cf250a0427e9a0e439b75
+oid sha256:712623d8718acf932a697c6a2fdaff6e6224128cb9406b1abb111776117aa39f
 size 80881450