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/a40fb77895c3bbd2_train_data.json
 model_type: AutoModelForCausalLM
@@ -90,7 +90,7 @@ wandb_name: e8e5f0fb-dd8b-49b4-81d6-934ffc0c9f10
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: e8e5f0fb-dd8b-49b4-81d6-934ffc0c9f10
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -101,6 +101,8 @@ xformers_attention: null
 # 8156638e-3410-43a8-b5a3-22180e93b934
 
 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: 1.2174
 
 ## 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        | 0.0024 | 1    | 1.4359          |
+| 5.4448        | 0.0240 | 10   | 1.3160          |
+| 5.235         | 0.0480 | 20   | 1.2547          |
+| 5.1044        | 0.0720 | 30   | 1.2333          |
+| 4.9788        | 0.0960 | 40   | 1.2202          |
+| 4.637         | 0.1200 | 50   | 1.2174          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:721a562b3e35c083db82390160ebef815c7043b77fc806bab32c9dbe06f1ecad
+oid sha256:0f3ee182082a4a3b8aaf94147f3672cb2df5dba3b3d7de1222eac6ce744435ac
 size 84047370