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/c4c19b6ea43684de_train_data.json
 model_type: AutoModelForCausalLM
@@ -92,7 +92,7 @@ wandb_name: d25a3c97-81bd-4006-b8c2-72fefd1150ec
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: d25a3c97-81bd-4006-b8c2-72fefd1150ec
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -103,6 +103,8 @@ xformers_attention: null
 # d25a3c97-81bd-4006-b8c2-72fefd1150ec
 
 This model is a fine-tuned version of [tiiuae/falcon-rw-1b](https://huggingface.co/tiiuae/falcon-rw-1b) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.7085
 
 ## Model description
 
@@ -129,14 +131,19 @@ 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        | 0.0017 | 1    | 8.2931          |
+| 29.2695       | 0.0168 | 10   | 5.2296          |
+| 7.6595        | 0.0337 | 20   | 1.0594          |
+| 3.2348        | 0.0505 | 30   | 0.7455          |
+| 2.8342        | 0.0674 | 40   | 0.7106          |
+| 2.9617        | 0.0842 | 50   | 0.7085          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f288cc4c7a897d3a3d7d2fb1fb96d98300b883e501de2147ed98d38390fce80c
+oid sha256:a9b20672024ea84bb38b0839de16bcc2bccc4913a6ee928e15988d8f7057c474
 size 25236362