End of training

README.md

@@ -64,7 +64,7 @@ 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/244a2945e2373286_train_data.json
 model_type: AutoModelForCausalLM
@@ -89,7 +89,7 @@ wandb_name: 5e3967ba-d17c-4d93-91ec-a23620abb5dc
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 5e3967ba-d17c-4d93-91ec-a23620abb5dc
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -100,6 +100,8 @@ xformers_attention: null
 # b9dfec1b-3f21-42cc-b6a6-17ad73fb75c7
 
 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: 11.9336
 
 ## Model description
 
@@ -126,14 +128,19 @@ 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    | 11.9376         |
+| 11.9363       | 0.0007 | 10   | 11.9370         |
+| 11.9374       | 0.0014 | 20   | 11.9356         |
+| 11.9319       | 0.0022 | 30   | 11.9343         |
+| 11.9326       | 0.0029 | 40   | 11.9337         |
+| 11.9338       | 0.0036 | 50   | 11.9336         |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3ced61525329edcee42311e542e8aaf8b065891566480e0f8479c50708800fa6
+oid sha256:eed320bb43cb96058d94627762d0d7eab58f75ec8dd92d4e55b6ede8863f8722
 size 388266