End of training

README.md

@@ -59,7 +59,7 @@ 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/71dcf982b77fb12b_train_data.json
 model_type: AutoModelForCausalLM
@@ -84,7 +84,7 @@ wandb_name: b4ebaae4-89cb-407e-9620-506121fb94ad
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: b4ebaae4-89cb-407e-9620-506121fb94ad
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -95,6 +95,8 @@ xformers_attention: null
 # b4ebaae4-89cb-407e-9620-506121fb94ad
 
 This model is a fine-tuned version of [sethuiyer/Medichat-Llama3-8B](https://huggingface.co/sethuiyer/Medichat-Llama3-8B) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 2.8090
 
 ## Model description
 
@@ -121,14 +123,19 @@ 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        | 0.0000 | 1    | 3.5952          |
+| 3.3319        | 0.0003 | 10   | 3.3436          |
+| 3.088         | 0.0006 | 20   | 2.9063          |
+| 2.7306        | 0.0009 | 30   | 2.8463          |
+| 2.7671        | 0.0013 | 40   | 2.8151          |
+| 2.5389        | 0.0016 | 50   | 2.8090          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8dc61b8f36ddf79ee5984afb039702d06d27fb8d92a0e145dd09d6c7803010a8
+oid sha256:0122955511e490a94c2cca497604e697b7d1e658eebe8e8c8abfae35cb20b374
 size 84047370