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/e3c32199351278e4_train_data.json
 model_type: AutoModelForCausalLM
@@ -91,7 +91,7 @@ wandb_name: aba65c6b-5ce7-48c9-973e-6598133ba9d0
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: aba65c6b-5ce7-48c9-973e-6598133ba9d0
-warmup_steps: 1
+warmup_steps: 10
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,6 +102,8 @@ xformers_attention: null
 # d92a2c65-942d-43a5-bcfd-e3529cdf31bd
 
 This model is a fine-tuned version of [OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5](https://huggingface.co/OpenAssistant/oasst-sft-4-pythia-12b-epoch-3.5) on the None dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.7428
 
 ## 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        | 0.0025 | 1    | 0.8764          |
+| 3.4396        | 0.0249 | 10   | 0.8358          |
+| 3.2492        | 0.0498 | 20   | 0.7855          |
+| 2.9669        | 0.0747 | 30   | 0.7551          |
+| 3.0397        | 0.0996 | 40   | 0.7451          |
+| 3.0665        | 0.1245 | 50   | 0.7428          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e8e571b76693ffeb9d2a63365a47f7ed13973c021703f413255f010ee6a5b9ba
+oid sha256:032f5c958cfdca95bf21f9d9468639c943f92a12614873ce836e5f9bd64e3d66
 size 94477642