Model save

README.md

@@ -16,7 +16,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [HuggingFaceTB/SmolLM2-135M-Instruct](https://huggingface.co/HuggingFaceTB/SmolLM2-135M-Instruct) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9152
+- Loss: 0.7844
 
 ## Model description
 
@@ -36,25 +36,28 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 0.0003
-- train_batch_size: 4
-- eval_batch_size: 4
+- train_batch_size: 8
+- eval_batch_size: 8
 - seed: 2024
 - distributed_type: multi-GPU
 - num_devices: 8
 - gradient_accumulation_steps: 8
-- total_train_batch_size: 256
-- total_eval_batch_size: 32
+- total_train_batch_size: 512
+- total_eval_batch_size: 64
 - optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_ratio: 0.1
-- training_steps: 100
+- num_epochs: 2
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss |
-|:-------------:|:-----:|:----:|:---------------:|
-| 1.1023        | 12.5  | 50   | 0.9144          |
-| 1.0625        | 25.0  | 100  | 0.9152          |
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 0.9921        | 0.3853 | 500  | 0.8058          |
+| 0.9685        | 0.7706 | 1000 | 0.7897          |
+| 1.0532        | 1.1558 | 1500 | 0.7858          |
+| 1.0206        | 1.5411 | 2000 | 0.7847          |
+| 1.0418        | 1.9264 | 2500 | 0.7844          |
 
 
 ### Framework versions

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3fb1c4c47a76e7ab40951b0bb2709a838b1aaf4924e205efd727bb56db027e3c
+oid sha256:cdca2476fdf69458c6ad0fc334a14db86907ab84020b2b6f31a21762f3e8b9fa
 size 269060552