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/a176eebac3e98bce_train_data.json
 model_type: AutoModelForCausalLM
@@ -93,7 +93,7 @@ wandb_name: f9eb2fff-bae0-49b0-89c5-624fc71c75d0
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: f9eb2fff-bae0-49b0-89c5-624fc71c75d0
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -104,6 +104,8 @@ xformers_attention: null
 # f9eb2fff-bae0-49b0-89c5-624fc71c75d0
 
 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: 1.7841
 
 ## 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        | 0.0001 | 1    | 5.3159          |
+| 4.9677        | 0.0010 | 10   | 4.6847          |
+| 3.6173        | 0.0020 | 20   | 3.2087          |
+| 2.5896        | 0.0030 | 30   | 2.2779          |
+| 1.9985        | 0.0039 | 40   | 1.8193          |
+| 1.8765        | 0.0049 | 50   | 1.7841          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:018c78896a88c98a016f5dc1803ba57eac30fad1a3da9638feb282f9b149fe1f
+oid sha256:2f3117b3c8b277c3db221f3fd6aac85b8aca1502cefe46c4a8f7f5161e3c23b9
 size 6843306