Upload folder using huggingface_hub

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bcc99ddb69f6afd56f9b1b95b8e6aec14a5ee954e8ff83c4b2fb885211bebef2
+oid sha256:db3ea32b05a90d614dfc081fce01b86b6256ac7d62d799e1d43c410f50ae3e67
 size 988097824

training_artifacts/README.md

@@ -12,5 +12,5 @@ This directory contains the training configuration and logs for this model.
 ## Job Information
 
 - Job Name: lf_torch_test__interactive
-- Timestamp: 2025-10-22 21:26:55 UTC
+- Timestamp: 2025-10-23 00:36:20 UTC
 - Execution Mode: Local

training_artifacts/hydra_config.yaml

@@ -167,7 +167,7 @@ lora:
   dropout: 0.05
   target: all
 dataset:
-  name: my_custom_sft
+  name: my_custom_sft12
   dir: null
   info_json: null
   template: default
@@ -187,6 +187,8 @@ dataset:
     content: content
     user: user
     assistant: assistant
+  tokenized_path: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/tokenized/my_custom_sft12
+  data_shared_file_system: true
 output:
   experiment_dir: ./experiments
 wandb:
@@ -213,4 +215,6 @@ training:
   per_device_train_batch_size: 1
   gradient_accumulation_steps: 1
   gradient_checkpointing: true
+  preprocessing_num_workers: 16
+  overwrite_cache: true
 merge: {}

training_artifacts/train_config.yaml

@@ -11,18 +11,19 @@ overwrite_output_dir: true
 per_device_train_batch_size: 1
 gradient_accumulation_steps: 1
 gradient_checkpointing: true
+preprocessing_num_workers: 16
+overwrite_cache: true
 model_name_or_path: Qwen/Qwen2.5-0.5B
 finetuning_type: lora
-dataset: my_custom_sft
+dataset: my_custom_sft12
 dataset_dir: /scratch/zrs2020/LlamaFactoryHelper/LLaMA-Factory/data
 template: default
 cutoff_len: 8096
 val_size: 0.1
+tokenized_path: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/tokenized/my_custom_sft12
+data_shared_file_system: true
 lora_rank: 8
 lora_alpha: 16
 lora_dropout: 0.05
 lora_target: all
 output_dir: /scratch/zrs2020/LlamaFactoryHelper/experiments/lf_torch_test__interactive/checkpoints
-preprocessing_num_workers: 1
-disable_tqdm: false
-overwrite_cache: true