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/9d35c252a83a77ae_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: e5161a80-d0e8-494e-bcf8-dc21c6737f76
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: e5161a80-d0e8-494e-bcf8-dc21c6737f76
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # e5161a80-d0e8-494e-bcf8-dc21c6737f76
 
 This model is a fine-tuned version of [unsloth/llama-2-7b-chat](https://huggingface.co/unsloth/llama-2-7b-chat) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0321
 
 ## Model description
 
@@ -128,14 +130,19 @@ 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    | 4.4015          |
+| 3.4607        | 0.0019 | 10   | 1.6497          |
+| 0.0126        | 0.0038 | 20   | 0.0455          |
+| 0.0           | 0.0057 | 30   | 0.0359          |
+| 0.0003        | 0.0076 | 40   | 0.0326          |
+| 0.0           | 0.0095 | 50   | 0.0321          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:70647b5a24436a26c42093a86e80ae3eacf294cf5c360563db12d8c82a21688a
+oid sha256:454a4333897642161e9bca66d3f21cb1bb171891ebcaa9b1b4e20f93bdb1899a
 size 80115210