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/2873280c1551a4bb_train_data.json
 model_type: AutoModelForCausalLM
@@ -93,7 +93,7 @@ wandb_name: d8427ff2-5f59-4882-b0ec-6b02abb5a270
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: d8427ff2-5f59-4882-b0ec-6b02abb5a270
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -104,6 +104,8 @@ xformers_attention: null
 # d8427ff2-5f59-4882-b0ec-6b02abb5a270
 
 This model is a fine-tuned version of [tiiuae/falcon-7b](https://huggingface.co/tiiuae/falcon-7b) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.2198
 
 ## 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    | 10.3461         |
+| 38.9418       | 0.0016 | 10   | 7.5474          |
+| 7.9997        | 0.0032 | 20   | 0.9896          |
+| 1.5443        | 0.0048 | 30   | 0.3319          |
+| 0.8458        | 0.0064 | 40   | 0.2406          |
+| 0.9225        | 0.0079 | 50   | 0.2198          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:01134af21740a267f11026e9c0a625577163f24c83d97f399786ae24c0d3cd52
+oid sha256:ff8bb05f8aa85c334a4e4bebf4073ecc06068a4139f4990c43ff22d6c13ead63
 size 65367946