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/08f60fa8472c9c67_train_data.json
 model_type: AutoModelForCausalLM
@@ -92,7 +92,7 @@ wandb_name: 7cf1683a-3eda-492d-91cd-27fe5e45e5ca
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 7cf1683a-3eda-492d-91cd-27fe5e45e5ca
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -103,6 +103,8 @@ xformers_attention: null
 # 1762c93b-4ab4-4708-a3f0-48302bc348da
 
 This model is a fine-tuned version of [katuni4ka/tiny-random-falcon-40b](https://huggingface.co/katuni4ka/tiny-random-falcon-40b) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 11.0784
 
 ## Model description
 
@@ -129,14 +131,18 @@ 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: 5
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
 | No log        | 0.2105 | 1    | 11.0850         |
+| No log        | 0.4211 | 2    | 11.0850         |
+| No log        | 0.6316 | 3    | 11.0845         |
+| No log        | 0.8421 | 4    | 11.0812         |
+| 44.3307       | 1.1579 | 5    | 11.0784         |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5d00607d16e8068d3241b392ed21875d4ea0814fe40481b2331f4a7f4a0fd23c
+oid sha256:7d272aa695115e66451339a30d4ac0420b5249cc1836a20a067bce4c55806b41
 size 129182