End of training

README.md

@@ -67,7 +67,7 @@ lora_target_linear: true
 lr_scheduler: cosine
 max_memory:
   0: 45GB
-max_steps: 1
+max_steps: 50
 micro_batch_size: 2
 mlflow_experiment_name: /tmp/8421daf5d3e18daa_train_data.json
 model_type: AutoModelForCausalLM
@@ -94,7 +94,7 @@ wandb_name: bb738fc3-1624-4a5e-a153-4f5e543b3e8c
 wandb_project: Gradients-On-Demand
 wandb_run: unknown
 wandb_runid: bb738fc3-1624-4a5e-a153-4f5e543b3e8c
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -105,6 +105,8 @@ xformers_attention: null
 # ea39adaf-acf3-48a1-b995-7cad860ed8a2
 
 This model is a fine-tuned version of [fxmarty/really-tiny-falcon-testing](https://huggingface.co/fxmarty/really-tiny-falcon-testing) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 11.0854
 
 ## Model description
 
@@ -131,14 +133,19 @@ 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        | 0.0027 | 1    | 11.0936         |
+| 44.3673       | 0.0266 | 10   | 11.0917         |
+| 44.3899       | 0.0533 | 20   | 11.0896         |
+| 44.3692       | 0.0799 | 30   | 11.0872         |
+| 44.3546       | 0.1065 | 40   | 11.0856         |
+| 44.3605       | 0.1332 | 50   | 11.0854         |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a0934f48d1cdc80e5a80cf0be4abff54e6062838693a7aa1910c33d4c5984957
+oid sha256:3f381fea8ac38145d8e84f43f2e7b1e28aecd34b60f7522e99a769cb5bc901c5
 size 37022