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: 8
 mlflow_experiment_name: /tmp/c02fe2e5c55fb819_train_data.json
 model_type: AutoModelForCausalLM
@@ -93,7 +93,7 @@ wandb_name: 580bc260-eb0c-4964-a47a-bc876e8cb2d3
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 580bc260-eb0c-4964-a47a-bc876e8cb2d3
-warmup_steps: 1
+warmup_steps: 2
 weight_decay: 0.0
 xformers_attention: null
 
@@ -104,6 +104,8 @@ xformers_attention: null
 # 580bc260-eb0c-4964-a47a-bc876e8cb2d3
 
 This model is a fine-tuned version of [echarlaix/tiny-random-mistral](https://huggingface.co/echarlaix/tiny-random-mistral) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 10.3737
 
 ## Model description
 
@@ -131,13 +133,18 @@ The following hyperparameters were used during training:
 - 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
+- training_steps: 50
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
 | No log        | 0.0022 | 1    | 10.3768         |
+| 166.0306      | 0.0222 | 10   | 10.3759         |
+| 166.0063      | 0.0445 | 20   | 10.3748         |
+| 165.9956      | 0.0667 | 30   | 10.3741         |
+| 165.981       | 0.0889 | 40   | 10.3737         |
+| 165.984       | 0.1111 | 50   | 10.3737         |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:577c2b2c87396dfc6e50a40ebf212b2b0a9bfbade834837415c141826d78e015
+oid sha256:bf76fef315f82dd4fee83a4c6fd1533e3045fcaa021bd2e2c20602b13782c669
 size 37506