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/92f639a945928a72_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: 242e80a5-2e54-45b1-9f78-864413afaf65
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 242e80a5-2e54-45b1-9f78-864413afaf65
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # 4c074535-f3ee-53c6-8807-db4952db55c5
 
 This model is a fine-tuned version of [Qwen/Qwen1.5-1.8B](https://huggingface.co/Qwen/Qwen1.5-1.8B) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0946
 
 ## 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        | 0.0001 | 1    | 4.2746          |
+| 2.5334        | 0.0007 | 10   | 2.8178          |
+| 0.7616        | 0.0014 | 20   | 0.4284          |
+| 0.2611        | 0.0021 | 30   | 0.1142          |
+| 0.1585        | 0.0027 | 40   | 0.0976          |
+| 0.1505        | 0.0034 | 50   | 0.0946          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8e6ecd9670a5651c295a22ab284818cefb3c9cc01a3582ee3d4732a3fb292437
+oid sha256:2fb4440e5d47a08a137b28f5cc1e378023d9ac65df3a323be95f18aa3c91e961
 size 30103498