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/e0aeaec1d3e659ac_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: baf41aeb-d82b-4e1a-b4a6-4158be1196e9
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: baf41aeb-d82b-4e1a-b4a6-4158be1196e9
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # fa133dca-99cf-4d7b-aa58-5588aa5e6b53
 
 This model is a fine-tuned version of [aisingapore/llama3-8b-cpt-sea-lionv2.1-instruct](https://huggingface.co/aisingapore/llama3-8b-cpt-sea-lionv2.1-instruct) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0091
 
 ## 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        | 0.0006 | 1    | 0.6192          |
+| 0.4819        | 0.0064 | 10   | 0.2438          |
+| 0.0513        | 0.0127 | 20   | 0.0443          |
+| 0.0232        | 0.0191 | 30   | 0.0134          |
+| 0.0097        | 0.0254 | 40   | 0.0101          |
+| 0.0071        | 0.0318 | 50   | 0.0091          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6060130cadf91d8f974ec44082e8408aed7356e29f61c1dbce6cda9cd38e68c2
+oid sha256:7fade37b483f9be65dca2ab38c0f8751ad18f217b278c29ff54464da6e25513b
 size 84047370