End of training

README.md

@@ -65,7 +65,7 @@ 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/2297fbeb8849d966_train_data.json
 model_type: AutoModelForCausalLM
@@ -90,7 +90,7 @@ wandb_name: 51c819d4-36da-4ea1-8617-f56fbac283ee
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 51c819d4-36da-4ea1-8617-f56fbac283ee
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -101,6 +101,8 @@ xformers_attention: null
 # dc30e44b-f60c-5ed0-789c-97031c59063e
 
 This model is a fine-tuned version of [unsloth/llama-3-8b-Instruct](https://huggingface.co/unsloth/llama-3-8b-Instruct) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.3501
 
 ## Model description
 
@@ -127,14 +129,19 @@ 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        | 0.0004 | 1    | 1.3328          |
+| 1.1261        | 0.0043 | 10   | 0.9592          |
+| 0.5037        | 0.0085 | 20   | 0.4851          |
+| 0.3204        | 0.0128 | 30   | 0.3878          |
+| 0.3335        | 0.0171 | 40   | 0.3559          |
+| 0.4512        | 0.0213 | 50   | 0.3501          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5056bd4d3d997cfa1a70e9205a38315539113eb8f6d76d9a6538018946133ea8
+oid sha256:55d005cade5bdf52a276c9ad17606a3bf9642fe7493e2f2c21b421d91c89a662
 size 84047370