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/899a7ca23f86f3d7_train_data.json
 model_type: AutoModelForCausalLM
@@ -93,7 +93,7 @@ wandb_name: b043e58c-eadc-4a82-808d-5cb5ff10094c
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: b043e58c-eadc-4a82-808d-5cb5ff10094c
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -104,6 +104,8 @@ xformers_attention: null
 # b043e58c-eadc-4a82-808d-5cb5ff10094c
 
 This model is a fine-tuned version of [JackFram/llama-160m](https://huggingface.co/JackFram/llama-160m) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.8174
 
 ## Model description
 
@@ -130,14 +132,19 @@ 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    | 3.3041          |
+| 3.2666        | 0.0025 | 10   | 3.2536          |
+| 3.0345        | 0.0050 | 20   | 3.0423          |
+| 2.9029        | 0.0074 | 30   | 2.8979          |
+| 2.7674        | 0.0099 | 40   | 2.8300          |
+| 2.8546        | 0.0124 | 50   | 2.8174          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:85749ae012af23dc3ae09a97c25198f667dee594a33a331f829aaefada8de16d
+oid sha256:594004dd513adfc5cd8d31b6990d62a195bee4d31192302ae4d3877eaf4c3a75
 size 6843306