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/5dbfb03345615677_train_data.json
 model_type: AutoModelForCausalLM
@@ -90,7 +90,7 @@ wandb_name: e2c36b69-018d-47fa-9a39-f9ee7db13418
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: e2c36b69-018d-47fa-9a39-f9ee7db13418
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -101,6 +101,8 @@ xformers_attention: null
 # e2c36b69-018d-47fa-9a39-f9ee7db13418
 
 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: 0.0414
 
 ## 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        | 0.0004 | 1    | 1.4710          |
+| 4.5091        | 0.0040 | 10   | 0.7016          |
+| 0.8607        | 0.0081 | 20   | 0.1132          |
+| 0.3276        | 0.0121 | 30   | 0.0690          |
+| 0.1905        | 0.0161 | 40   | 0.0450          |
+| 0.1571        | 0.0202 | 50   | 0.0414          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:43199c0f85cf93444a5a1fb75a9951c0cdc4c304962bca230a8f9149224fd443
+oid sha256:dfbb46a2f4b42030bc6a1fd82dd2bfe8e54bc7a2c763abead237091aaaf16c9b
 size 84047370