End of training

README.md

@@ -46,7 +46,7 @@ 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/ade6d149-5430-4614-9b8f-0bebe06ee174
@@ -57,7 +57,7 @@ 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
@@ -66,7 +66,7 @@ lora_r: 8
 lora_target_linear: true
 lr_scheduler: cosine
 max_steps: 1
-micro_batch_size: 8
+micro_batch_size: 2
 mlflow_experiment_name: /tmp/b3eb0292cb32b6ac_train_data.json
 model_type: AutoModelForCausalLM
 num_epochs: 1
@@ -120,11 +120,11 @@ 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
@@ -134,7 +134,7 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss |
 |:-------------:|:------:|:----:|:---------------:|
-| No log        | 0.0349 | 1    | 7.9264          |
+| No log        | 0.0022 | 1    | 7.9420          |
 
 
 ### Framework versions

adapter_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fa2ab8b537a6231fa3e2cfe939219a4e5882aa339f3090025c6f06e7b4a37880
+oid sha256:29e1774b592bd0e3ee0f02c3712168b8a8735acbbe19ed67f5746900b52a2214
 size 100149034