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: 2
 mlflow_experiment_name: /tmp/40388247a6d648af_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 55f164e0-712d-4139-bc38-6c9689b659ea
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 55f164e0-712d-4139-bc38-6c9689b659ea
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # cde94493-0b03-61c4-7173-1f544479f55d
 
 This model is a fine-tuned version of [unsloth/mistral-7b-instruct-v0.2](https://huggingface.co/unsloth/mistral-7b-instruct-v0.2) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.3294
 
 ## Model description
 
@@ -128,14 +130,19 @@ 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    | 0.4858          |
+| 1.7327        | 0.0004 | 10   | 0.3959          |
+| 1.356         | 0.0008 | 20   | 0.3506          |
+| 1.2296        | 0.0011 | 30   | 0.3363          |
+| 1.4278        | 0.0015 | 40   | 0.3304          |
+| 1.2683        | 0.0019 | 50   | 0.3294          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d45605d0b9cd5e7fef88653b9aa490417a99a27ebf6d358ee5b9b09af3f4029c
+oid sha256:d1ffba5e5efb09e88191f4d4b5961b74f0a056bca1d6bc6cdc4a40407e0a4b53
 size 84047370