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/ddde6542fd09fb6d_train_data.json
 model_type: AutoModelForCausalLM
@@ -93,7 +93,7 @@ wandb_name: 599737b6-a5d2-4a15-bc69-d7e37d8ba0dd
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 599737b6-a5d2-4a15-bc69-d7e37d8ba0dd
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -104,6 +104,8 @@ xformers_attention: null
 # f406aef8-6f43-48c9-a2a0-90d625d447c1
 
 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: 0.3328
 
 ## 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    | 1.2216          |
+| 1.1951        | 0.0018 | 10   | 1.1232          |
+| 0.8239        | 0.0037 | 20   | 0.7094          |
+| 0.4812        | 0.0055 | 30   | 0.4219          |
+| 0.3619        | 0.0074 | 40   | 0.3449          |
+| 0.3192        | 0.0092 | 50   | 0.3328          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c5e0f6c6c8bd1ce3e032dfa9e268612cf86f967f3a945936a0a11edd4b45a7ed
+oid sha256:64bd1c7c8d1d7668e7c30034fcf9bb5c658b371dc590bcf09b547f3ed7211ee0
 size 50982842