End of training

README.md

@@ -65,7 +65,7 @@ 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/d43c095226bab7a0_train_data.json
 model_type: AutoModelForCausalLM
@@ -90,7 +90,7 @@ wandb_name: 03359f4a-dd6a-4c9a-ba1c-156f3c868b27
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 03359f4a-dd6a-4c9a-ba1c-156f3c868b27
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -101,6 +101,8 @@ xformers_attention: null
 # fbe64908-82f4-3ba4-3aac-abe2b32184a8
 
 This model is a fine-tuned version of [jhflow/mistral7b-lora-multi-turn-v2](https://huggingface.co/jhflow/mistral7b-lora-multi-turn-v2) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.1072
 
 ## Model description
 
@@ -127,14 +129,19 @@ 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    | 0.8266          |
+| 1.898         | 0.0005 | 10   | 0.3051          |
+| 0.7702        | 0.0010 | 20   | 0.1908          |
+| 0.3998        | 0.0015 | 30   | 0.1172          |
+| 0.3765        | 0.0020 | 40   | 0.1112          |
+| 0.2073        | 0.0025 | 50   | 0.1072          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6736bf3e4fc10b76213e3d79e1a5612574acd7732418c260a8c649de7012207c
+oid sha256:55c490a937376bc4079f3b9526c5f82c1a554f4f794375ad24c1fb42bd67c7a3
 size 84047370