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/511bc7520306da8a_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 9d3d6c1c-01e3-47b5-8f25-e2bdb653d048
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 9d3d6c1c-01e3-47b5-8f25-e2bdb653d048
-warmup_steps: 1
+warmup_steps: 2
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 9d3d6c1c-01e3-47b5-8f25-e2bdb653d048
 
 This model is a fine-tuned version of [TinyLlama/TinyLlama-1.1B-Chat-v0.6](https://huggingface.co/TinyLlama/TinyLlama-1.1B-Chat-v0.6) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.5400
 
 ## 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        | 0.0051 | 1    | 0.6309          |
+| 0.6061        | 0.0511 | 10   | 0.6027          |
+| 0.5841        | 0.1021 | 20   | 0.5628          |
+| 0.5459        | 0.1532 | 30   | 0.5474          |
+| 0.5444        | 0.2042 | 40   | 0.5410          |
+| 0.5309        | 0.2553 | 50   | 0.5400          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5487c9f1f2d0e51f2212a835b5a0e44194606e58eb76d4f2e5126b076c3ef923
+oid sha256:ad33962941d36aca7856cc2eaea138b62c367eff8fe3b2e5eb227d967e1eb70e
 size 25342042