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: 8
 mlflow_experiment_name: /tmp/dcb983c631a7d8d7_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: a45cdd93-5d74-41cd-8f7d-d7a30c644afe
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: a45cdd93-5d74-41cd-8f7d-d7a30c644afe
-warmup_steps: 1
+warmup_steps: 2
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # a45cdd93-5d74-41cd-8f7d-d7a30c644afe
 
 This model is a fine-tuned version of [fxmarty/tiny-dummy-qwen2](https://huggingface.co/fxmarty/tiny-dummy-qwen2) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 11.9321
 
 ## Model description
 
@@ -129,13 +131,18 @@ The following hyperparameters were used during training:
 - 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
+- training_steps: 50
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
 | No log        | 0.0066 | 1    | 11.9322         |
+| 11.9322       | 0.0663 | 10   | 11.9322         |
+| 11.9321       | 0.1327 | 20   | 11.9321         |
+| 11.9322       | 0.1990 | 30   | 11.9321         |
+| 11.932        | 0.2653 | 40   | 11.9321         |
+| 11.9319       | 0.3317 | 50   | 11.9321         |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9b893de670c04d2dbdd5a81603259b146ede91fdde83db321d11a018501a7ae0
+oid sha256:1a3b8bf2878a32476ca8a4760395df78c0665d2f15596eb4a6504d99d9b50124
 size 21378