End of training

README.md

@@ -47,7 +47,7 @@ flash_attention: false
 fp16: null
 fsdp: null
 fsdp_config: null
-gradient_accumulation_steps: 16
+gradient_accumulation_steps: 4
 gradient_checkpointing: false
 group_by_length: false
 hub_model_id: tuanna08go/85127c1a-1e48-44f0-9d81-717f87454d94
@@ -58,7 +58,7 @@ learning_rate: 0.0001
 load_in_4bit: false
 load_in_8bit: false
 local_rank: null
-logging_steps: 10
+logging_steps: 5
 lora_alpha: 16
 lora_dropout: 0.05
 lora_fan_in_fan_out: null
@@ -66,8 +66,8 @@ lora_model_dir: null
 lora_r: 8
 lora_target_linear: true
 lr_scheduler: cosine
-max_steps: 50
-micro_batch_size: 8
+max_steps: 1
+micro_batch_size: 2
 mlflow_experiment_name: /tmp/160c5b3bb9da3f56_train_data.json
 model_type: AutoModelForCausalLM
 num_epochs: 1
@@ -91,7 +91,7 @@ wandb_name: 85127c1a-1e48-44f0-9d81-717f87454d94
 wandb_project: Gradients-On-Demand
 wandb_run: your_name
 wandb_runid: 85127c1a-1e48-44f0-9d81-717f87454d94
-warmup_steps: 2
+warmup_steps: 1
 weight_decay: 0.0
 xformers_attention: null
 
@@ -102,8 +102,6 @@ xformers_attention: null
 # 85127c1a-1e48-44f0-9d81-717f87454d94
 
 This model is a fine-tuned version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-Instruct) on the None dataset.
-It achieves the following results on the evaluation set:
-- Loss: 2.5323
 
 ## Model description
 
@@ -123,26 +121,21 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 0.0001
-- train_batch_size: 8
-- eval_batch_size: 8
+- train_batch_size: 2
+- eval_batch_size: 2
 - seed: 42
-- gradient_accumulation_steps: 16
-- total_train_batch_size: 128
+- gradient_accumulation_steps: 4
+- 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: 50
+- training_steps: 1
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| No log        | 0.0050 | 1    | 2.7468          |
-| 2.7375        | 0.0500 | 10   | 2.6682          |
-| 2.6298        | 0.0999 | 20   | 2.6001          |
-| 2.5657        | 0.1499 | 30   | 2.5527          |
-| 2.555         | 0.1999 | 40   | 2.5356          |
-| 2.548         | 0.2498 | 50   | 2.5323          |
+| No log        | 0.0003 | 1    | nan             |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0d6adbd64856facc2276709eaa1eff14c737041e1e1e5580ff819784f3a1eefb
+oid sha256:5e9db931ceacf3857278160422aa613bf54d8250ae63b68e322736f3bb5ec808
 size 60048618