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/a0c669deb67c7ce4_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 8cb70d03-e02b-40be-b363-2d9ed3b3d9ff
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 8cb70d03-e02b-40be-b363-2d9ed3b3d9ff
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 8cb70d03-e02b-40be-b363-2d9ed3b3d9ff
 
 This model is a fine-tuned version of [unsloth/Llama-3.2-3B](https://huggingface.co/unsloth/Llama-3.2-3B) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.5040
 
 ## 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    | 0.6357          |
+| 0.6734        | 0.0017 | 10   | 0.6111          |
+| 0.5134        | 0.0034 | 20   | 0.5298          |
+| 0.4767        | 0.0051 | 30   | 0.5106          |
+| 0.4668        | 0.0068 | 40   | 0.5052          |
+| 0.4794        | 0.0085 | 50   | 0.5040          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7d4f099c76bea68a8de1f29d0da466863770b02e7603df194597e733a0cd2761
+oid sha256:312108ade06ec7d11ef5b128cf553f11bb97a24dbbfaac2586984b510d949351
 size 48768810