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: 8
 mlflow_experiment_name: /tmp/c97e2c46781b1d51_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: dd45c3f5-5ed2-4527-a5a5-b4b580d79aaf
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: dd45c3f5-5ed2-4527-a5a5-b4b580d79aaf
-warmup_steps: 1
+warmup_steps: 2
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # dd45c3f5-5ed2-4527-a5a5-b4b580d79aaf
 
 This model is a fine-tuned version of [NousResearch/Meta-Llama-3-8B-Alternate-Tokenizer](https://huggingface.co/NousResearch/Meta-Llama-3-8B-Alternate-Tokenizer) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 1.4623
 
 ## Model description
 
@@ -129,13 +131,14 @@ The following hyperparameters were used during training:
 - 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
+- training_steps: 2
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
 | No log        | 0.5333 | 1    | 1.5013          |
+| No log        | 1.0667 | 2    | 1.4623          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:da9761cd7ac2c66dc51493eab1d170cd324837e41ee1a8fc8f106c97816adeb2
+oid sha256:8d9240701684e8e0f259c2eb0f10d98333a37549c3a2f7eb9841df0753653655
 size 84047370