Improve model card for QFFT model

README.md

@@ -1,7 +1,7 @@
 ---
+base_model: Qwen/Qwen2.5-7B-Instruct
 library_name: transformers
 license: other
-base_model: Qwen/Qwen2.5-7B-Instruct
 tags:
 - llama-factory
 - full
@@ -9,22 +9,23 @@ tags:
 model-index:
 - name: 7b_isntruct_pretrain
   results: []
+pipeline_tag: text-generation
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
 should probably proofread and complete it, then remove this comment. -->
 
-# 7b_isntruct_pretrain
+# 7b\_isntruct\_pretrain
 
-This model is a fine-tuned version of Qwen2.5-7B-Instruct on the S1_QFFT dataset.
+This model is a fine-tuned version of [Qwen2.5-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-7B-Instruct) on the S1\_QFFT dataset. It was trained using the Question-Free Fine-Tuning (QFFT) method, as described in the paper [QFFT, Question-Free Fine-Tuning for Adaptive Reasoning](https://huggingface.co/papers/2506.12860).
 
 ## Model description
 
-More information needed
+This model adapts a base LLM to leverage both concise and detailed reasoning patterns by fine-tuning it without input questions, learning directly from long chain-of-thought responses.
 
 ## Intended uses & limitations
 
-More information needed
+This model is intended for research purposes in the domain of adaptive reasoning and chain-of-thought generation. It may exhibit improved efficiency in generating reasoning steps compared to standard fine-tuning but may also inherit limitations from the base model and the fine-tuning dataset.
 
 ## Training and evaluation data
 
@@ -35,31 +36,29 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
-- train_batch_size: 1
-- eval_batch_size: 8
-- seed: 42
-- distributed_type: multi-GPU
-- num_devices: 8
-- gradient_accumulation_steps: 1
-- total_train_batch_size: 8
-- total_eval_batch_size: 32
-- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
-- lr_scheduler_type: cosine
-- lr_scheduler_warmup_ratio: 0.1
-- num_epochs: 6
-
-### Training results
 
+*   learning\_rate: 1e-05
+*   train\_batch\_size: 1
+*   eval\_batch\_size: 8
+*   seed: 42
+*   distributed\_type: multi-GPU
+*   num\_devices: 8
+*   gradient\_accumulation\_steps: 1
+*   total\_train\_batch\_size: 8
+*   total\_eval\_batch\_size: 32
+*   optimizer: Use OptimizerNames.ADAMW\_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer\_args=No additional optimizer arguments
+*   lr\_scheduler\_type: cosine
+*   lr\_scheduler\_warmup\_ratio: 0.1
+*   num\_epochs: 6
 
+### Training results
 
 ### Framework versions
 
-- Transformers 4.48.2
-- Pytorch 2.6.0+cu124
-- Datasets 3.2.0
-- Tokenizers 0.21.0
-
+*   Transformers 4.48.2
+*   Pytorch 2.6.0+cu124
+*   Datasets 3.2.0
+*   Tokenizers 0.21.0
 
 ## 📖 Citation
 
@@ -70,6 +69,6 @@ The following hyperparameters were used during training:
   year={2025},
   eprint={2506.12860},
   archivePrefix={arXiv},
-  primaryClass={cs.CL},
   url={https://arxiv.org/abs/2506.12860},
-}
+}
+```