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/02173ff2debcb2d2_train_data.json
 model_type: AutoModelForCausalLM
@@ -90,7 +90,7 @@ wandb_name: f7337394-4699-4976-b38a-2593e132e919
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: f7337394-4699-4976-b38a-2593e132e919
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -101,6 +101,8 @@ xformers_attention: null
 # e8364d36-6639-4040-8e95-ecf39065f6f1
 
 This model is a fine-tuned version of [katuni4ka/tiny-random-qwen1.5-moe](https://huggingface.co/katuni4ka/tiny-random-qwen1.5-moe) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 11.9207
 
 ## Model description
 
@@ -127,14 +129,19 @@ 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    | 11.9317         |
+| 11.9275       | 0.0011 | 10   | 11.9298         |
+| 11.9288       | 0.0021 | 20   | 11.9259         |
+| 11.9246       | 0.0032 | 30   | 11.9227         |
+| 11.9214       | 0.0043 | 40   | 11.9210         |
+| 11.9193       | 0.0054 | 50   | 11.9207         |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:21cfef886ea8fe0abbb91f136106864ba6405c40206af0f493e1b857f3b296b6
+oid sha256:c7ddbce18ae821080152b27a5ee63579174e0ceda8e6f8b68e7708174195a5f7
 size 388266