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/6c1d97e52ba4ac77_train_data.json
 model_type: AutoModelForCausalLM
@@ -92,7 +92,7 @@ wandb_name: 93d42f3d-bba5-4636-9ddb-bf55bfed231f
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 93d42f3d-bba5-4636-9ddb-bf55bfed231f
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -103,6 +103,8 @@ xformers_attention: null
 # 023c8247-0358-4f93-a426-4e390a73768b
 
 This model is a fine-tuned version of [openlm-research/open_llama_3b](https://huggingface.co/openlm-research/open_llama_3b) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.6327
 
 ## 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        | 0.0009 | 1    | 1.6661          |
+| 1.6251        | 0.0086 | 10   | 1.6641          |
+| 1.7312        | 0.0172 | 20   | 1.6527          |
+| 1.6974        | 0.0258 | 30   | 1.6401          |
+| 1.5712        | 0.0344 | 40   | 1.6339          |
+| 1.6575        | 0.0430 | 50   | 1.6327          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:edf5ed96f033762453fef5e3d12ccf5e7c870a31a41a6d6902c2ea587d7de35d
+oid sha256:aed356597179f8e6afa19a4cf95210f9b808dcf50af1762ec65e088547dd1902
 size 50982842