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/492f5e5b92086fbd_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 3cb3135a-7307-47c7-a8af-25d09093fc0c
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 3cb3135a-7307-47c7-a8af-25d09093fc0c
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 228a2858-09c4-41f1-8a36-8a0384d3cfed
 
 This model is a fine-tuned version of [unsloth/Llama-3.2-1B-Instruct](https://huggingface.co/unsloth/Llama-3.2-1B-Instruct) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.6076
 
 ## 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        | 0.0032 | 1    | 1.8799          |
+| 1.8398        | 0.0319 | 10   | 1.8053          |
+| 1.631         | 0.0638 | 20   | 1.6729          |
+| 1.6784        | 0.0958 | 30   | 1.6297          |
+| 1.6291        | 0.1277 | 40   | 1.6109          |
+| 1.6381        | 0.1596 | 50   | 1.6076          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3775ccbaf14ca4ae57c23c1dfb1475cb011514040a8c7874a83d08b1fa5923ef
+oid sha256:9245ac7fe4c44ee625cd2095da0e646f566ad27a2036f371a1d326d7231d9063
 size 22624970