Model save

README.md

@@ -0,0 +1,68 @@
+---
+library_name: transformers
+model_name: DeepSeek-R1-Distill-Qwen-1.5B-DPO
+tags:
+- generated_from_trainer
+- trl
+- dpo
+licence: license
+---
+
+# Model Card for DeepSeek-R1-Distill-Qwen-1.5B-DPO
+
+This model is a fine-tuned version of [None](https://huggingface.co/None).
+It has been trained using [TRL](https://github.com/huggingface/trl).
+
+## Quick start
+
+```python
+from transformers import pipeline
+
+question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
+generator = pipeline("text-generation", model="LuyiCui/DeepSeek-R1-Distill-Qwen-1.5B-DPO", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
+
+## Training procedure
+
+[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="150" height="24"/>](https://wandb.ai/cuiluyi/open-r1/runs/kfi2ffhj) 
+
+
+This model was trained with DPO, a method introduced in [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://huggingface.co/papers/2305.18290).
+
+### Framework versions
+
+- TRL: 0.17.0.dev0
+- Transformers: 4.51.2
+- Pytorch: 2.6.0
+- Datasets: 3.5.0
+- Tokenizers: 0.21.1
+
+## Citations
+
+Cite DPO as:
+
+```bibtex
+@inproceedings{rafailov2023direct,
+    title        = {{Direct Preference Optimization: Your Language Model is Secretly a Reward Model}},
+    author       = {Rafael Rafailov and Archit Sharma and Eric Mitchell and Christopher D. Manning and Stefano Ermon and Chelsea Finn},
+    year         = 2023,
+    booktitle    = {Advances in Neural Information Processing Systems 36: Annual Conference on Neural Information Processing Systems 2023, NeurIPS 2023, New Orleans, LA, USA, December 10 - 16, 2023},
+    url          = {http://papers.nips.cc/paper_files/paper/2023/hash/a85b405ed65c6477a4fe8302b5e06ce7-Abstract-Conference.html},
+    editor       = {Alice Oh and Tristan Naumann and Amir Globerson and Kate Saenko and Moritz Hardt and Sergey Levine},
+}
+```
+
+Cite TRL as:
+    
+```bibtex
+@misc{vonwerra2022trl,
+	title        = {{TRL: Transformer Reinforcement Learning}},
+	author       = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.9523809523809523,
+    "total_flos": 0.0,
+    "train_loss": 0.6895670831203461,
+    "train_runtime": 273.9395,
+    "train_samples": 4000,
+    "train_samples_per_second": 14.602,
+    "train_steps_per_second": 0.037
+}

generation_config.json

@@ -0,0 +1,9 @@
+{
+  "_from_model_config": true,
+  "bos_token_id": 151646,
+  "do_sample": true,
+  "eos_token_id": 151643,
+  "temperature": 0.6,
+  "top_p": 0.95,
+  "transformers_version": "4.51.2"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 0.9523809523809523,
+    "total_flos": 0.0,
+    "train_loss": 0.6895670831203461,
+    "train_runtime": 273.9395,
+    "train_samples": 4000,
+    "train_samples_per_second": 14.602,
+    "train_steps_per_second": 0.037
+}

trainer_state.json

@@ -0,0 +1,193 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9523809523809523,
+  "eval_steps": 500,
+  "global_step": 10,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 10.040461540222168,
+      "learning_rate": 0.0,
+      "logits/chosen": 0.5517578125,
+      "logits/rejected": 0.5625,
+      "logps/chosen": -384.5,
+      "logps/rejected": -438.0,
+      "loss": 0.6982,
+      "rewards/accuracies": 0.0833333358168602,
+      "rewards/chosen": -0.016666412353515625,
+      "rewards/margins": -0.009382247924804688,
+      "rewards/rejected": -0.00731658935546875,
+      "step": 1
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 8.76193904876709,
+      "learning_rate": 1e-06,
+      "logits/chosen": 0.54248046875,
+      "logits/rejected": 0.54052734375,
+      "logps/chosen": -347.5,
+      "logps/rejected": -419.5,
+      "loss": 0.6898,
+      "rewards/accuracies": 0.109375,
+      "rewards/chosen": -0.010042190551757812,
+      "rewards/margins": 0.00848388671875,
+      "rewards/rejected": -0.01854705810546875,
+      "step": 2
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 7.859212875366211,
+      "learning_rate": 9.728616793536587e-07,
+      "logits/chosen": 0.546875,
+      "logits/rejected": 0.5322265625,
+      "logps/chosen": -372.0,
+      "logps/rejected": -420.0,
+      "loss": 0.6895,
+      "rewards/accuracies": 0.0807291716337204,
+      "rewards/chosen": 0.007147789001464844,
+      "rewards/margins": 0.0082244873046875,
+      "rewards/rejected": -0.0010318756103515625,
+      "step": 3
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 8.137142181396484,
+      "learning_rate": 8.9471999940354e-07,
+      "logits/chosen": 0.50927734375,
+      "logits/rejected": 0.55322265625,
+      "logps/chosen": -374.5,
+      "logps/rejected": -434.5,
+      "loss": 0.6903,
+      "rewards/accuracies": 0.109375,
+      "rewards/chosen": -0.02203369140625,
+      "rewards/margins": 0.0063877105712890625,
+      "rewards/rejected": -0.028411865234375,
+      "step": 4
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 6.646923542022705,
+      "learning_rate": 7.75e-07,
+      "logits/chosen": 0.45166015625,
+      "logits/rejected": 0.46630859375,
+      "logps/chosen": -373.5,
+      "logps/rejected": -412.5,
+      "loss": 0.6938,
+      "rewards/accuracies": 0.1093750074505806,
+      "rewards/chosen": -0.02127838134765625,
+      "rewards/margins": 0.0022153854370117188,
+      "rewards/rejected": -0.0235137939453125,
+      "step": 5
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 8.178302764892578,
+      "learning_rate": 6.281416799501187e-07,
+      "logits/chosen": 0.533203125,
+      "logits/rejected": 0.544921875,
+      "logps/chosen": -387.0,
+      "logps/rejected": -442.5,
+      "loss": 0.6883,
+      "rewards/accuracies": 0.1223958358168602,
+      "rewards/chosen": -0.029693603515625,
+      "rewards/margins": 0.0110931396484375,
+      "rewards/rejected": -0.040740966796875,
+      "step": 6
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 9.006941795349121,
+      "learning_rate": 4.7185832004988133e-07,
+      "logits/chosen": 0.5537109375,
+      "logits/rejected": 0.5390625,
+      "logps/chosen": -368.0,
+      "logps/rejected": -418.0,
+      "loss": 0.6898,
+      "rewards/accuracies": 0.140625,
+      "rewards/chosen": -0.062744140625,
+      "rewards/margins": 0.009137153625488281,
+      "rewards/rejected": -0.0718994140625,
+      "step": 7
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 6.875649929046631,
+      "learning_rate": 3.250000000000001e-07,
+      "logits/chosen": 0.5703125,
+      "logits/rejected": 0.5859375,
+      "logps/chosen": -377.5,
+      "logps/rejected": -427.0,
+      "loss": 0.6864,
+      "rewards/accuracies": 0.1614583432674408,
+      "rewards/chosen": -0.06878662109375,
+      "rewards/margins": 0.01445770263671875,
+      "rewards/rejected": -0.083251953125,
+      "step": 8
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 7.995317459106445,
+      "learning_rate": 2.0528000059645995e-07,
+      "logits/chosen": 0.5068359375,
+      "logits/rejected": 0.544921875,
+      "logps/chosen": -388.0,
+      "logps/rejected": -447.0,
+      "loss": 0.6867,
+      "rewards/accuracies": 0.1822916716337204,
+      "rewards/chosen": -0.0706787109375,
+      "rewards/margins": 0.016702651977539062,
+      "rewards/rejected": -0.0875244140625,
+      "step": 9
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 9.462100982666016,
+      "learning_rate": 1.2713832064634125e-07,
+      "logits/chosen": 0.5576171875,
+      "logits/rejected": 0.560546875,
+      "logps/chosen": -362.0,
+      "logps/rejected": -437.0,
+      "loss": 0.6829,
+      "rewards/accuracies": 0.1953125,
+      "rewards/chosen": -0.0859375,
+      "rewards/margins": 0.022670745849609375,
+      "rewards/rejected": -0.108642578125,
+      "step": 10
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "step": 10,
+      "total_flos": 0.0,
+      "train_loss": 0.6895670831203461,
+      "train_runtime": 273.9395,
+      "train_samples_per_second": 14.602,
+      "train_steps_per_second": 0.037
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 10,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}