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/afa2f66839b5234c_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 71a25e11-de75-4ea7-9525-27c4e5d890cc
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 71a25e11-de75-4ea7-9525-27c4e5d890cc
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 804ce354-d5ef-c1bf-81c7-a3d393b3ed5d
 
 This model is a fine-tuned version of [unsloth/codegemma-7b](https://huggingface.co/unsloth/codegemma-7b) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.2592
 
 ## 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        | 0.0001 | 1    | 2.7244          |
+| 2.5449        | 0.0006 | 10   | 2.5042          |
+| 2.4901        | 0.0012 | 20   | 2.3432          |
+| 2.468         | 0.0019 | 30   | 2.2912          |
+| 2.2421        | 0.0025 | 40   | 2.2638          |
+| 2.2653        | 0.0031 | 50   | 2.2592          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b604470d8d5721dc28aa8053acc3e9d4ae09e0adf25638e098a1d8c8581a94a7
+oid sha256:464354594ae10d18124048977387d7bb09395b28beb2d9603dc44ea42a499da3
 size 100149034