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/ab80a554070cdc53_train_data.json
 model_type: AutoModelForCausalLM
@@ -93,7 +93,7 @@ wandb_name: 1b054327-a6b9-45e0-98ef-2d5fb503dfb8
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 1b054327-a6b9-45e0-98ef-2d5fb503dfb8
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -104,6 +104,8 @@ xformers_attention: null
 # b0af66e2-311e-58a8-f587-347aa494ddb6
 
 This model is a fine-tuned version of [JackFram/llama-160m](https://huggingface.co/JackFram/llama-160m) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 3.0715
 
 ## Model description
 
@@ -130,14 +132,19 @@ 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        | 0.0003 | 1    | 3.7050          |
+| 3.5779        | 0.0025 | 10   | 3.6573          |
+| 3.4933        | 0.0050 | 20   | 3.4188          |
+| 3.2524        | 0.0075 | 30   | 3.2014          |
+| 3.1295        | 0.0100 | 40   | 3.0912          |
+| 3.0719        | 0.0125 | 50   | 3.0715          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1ce47f6c0d02dd46ca33768f4e794d29ba35e4d60d9cdc952b5476c9b833ea10
+oid sha256:30a945c180bdc61b1abc15a73e6998b071e78cb843339f61bb52a4574a4869d6
 size 6843306