Model save

README.md

@@ -0,0 +1,58 @@
+---
+base_model: Qwen/Qwen2.5-0.5B-Instruct
+library_name: transformers
+model_name: Qwen2.5-0.5B-Open-R1-SFT
+tags:
+- generated_from_trainer
+- trl
+- sft
+licence: license
+---
+
+# Model Card for Qwen2.5-0.5B-Open-R1-SFT
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-0.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct).
+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="DGME/Qwen2.5-0.5B-Open-R1-SFT", 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/dgme-nanjing-university-of-aeronautics-and-astronautics/huggingface/runs/1cfo2dxw) 
+
+
+This model was trained with SFT.
+
+### Framework versions
+
+- TRL: 0.17.0.dev0
+- Transformers: 4.51.2
+- Pytorch: 2.6.0
+- Datasets: 3.3.2
+- Tokenizers: 0.21.0
+
+## Citations
+
+
+
+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 +1,8 @@
+{
+    "total_flos": 385299641270272.0,
+    "train_loss": 0.13468429481281954,
+    "train_runtime": 769.9529,
+    "train_samples": 7473,
+    "train_samples_per_second": 1.831,
+    "train_steps_per_second": 0.221
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 385299641270272.0,
+    "train_loss": 0.13468429481281954,
+    "train_runtime": 769.9529,
+    "train_samples": 7473,
+    "train_samples_per_second": 1.831,
+    "train_steps_per_second": 0.221
+}

trainer_state.json

@@ -0,0 +1,315 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 9.794326241134751,
+  "eval_steps": 500,
+  "global_step": 170,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.28368794326241137,
+      "grad_norm": 1.4403227431607122,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 0.5222,
+      "num_tokens": 634122.0,
+      "step": 5
+    },
+    {
+      "epoch": 0.5673758865248227,
+      "grad_norm": 1.051530473447262,
+      "learning_rate": 5e-05,
+      "loss": 0.4841,
+      "num_tokens": 1261436.0,
+      "step": 10
+    },
+    {
+      "epoch": 0.851063829787234,
+      "grad_norm": 0.5700469145307384,
+      "learning_rate": 4.989299698973645e-05,
+      "loss": 0.4265,
+      "num_tokens": 1900524.0,
+      "step": 15
+    },
+    {
+      "epoch": 1.1134751773049645,
+      "grad_norm": 0.4334982351318258,
+      "learning_rate": 4.957300570509739e-05,
+      "loss": 0.3977,
+      "num_tokens": 2498258.0,
+      "step": 20
+    },
+    {
+      "epoch": 1.397163120567376,
+      "grad_norm": 0.3413966057669926,
+      "learning_rate": 4.9043069704368563e-05,
+      "loss": 0.3787,
+      "num_tokens": 3127412.0,
+      "step": 25
+    },
+    {
+      "epoch": 1.6808510638297873,
+      "grad_norm": 0.30824886663026985,
+      "learning_rate": 4.83082294095344e-05,
+      "loss": 0.366,
+      "num_tokens": 3747537.0,
+      "step": 30
+    },
+    {
+      "epoch": 1.9645390070921986,
+      "grad_norm": 0.23393589833355252,
+      "learning_rate": 4.7375474164915813e-05,
+      "loss": 0.3629,
+      "num_tokens": 4390308.0,
+      "step": 35
+    },
+    {
+      "epoch": 2.226950354609929,
+      "grad_norm": 0.20565201227770782,
+      "learning_rate": 4.625367575886955e-05,
+      "loss": 0.3438,
+      "num_tokens": 4988042.0,
+      "step": 40
+    },
+    {
+      "epoch": 2.5106382978723403,
+      "grad_norm": 0.18576050641368966,
+      "learning_rate": 4.4953504040849445e-05,
+      "loss": 0.3479,
+      "num_tokens": 5633018.0,
+      "step": 45
+    },
+    {
+      "epoch": 2.794326241134752,
+      "grad_norm": 0.1668363092881809,
+      "learning_rate": 4.348732543642775e-05,
+      "loss": 0.337,
+      "num_tokens": 6264607.0,
+      "step": 50
+    },
+    {
+      "epoch": 3.0567375886524824,
+      "grad_norm": 0.2191426033169888,
+      "learning_rate": 4.1869085325538425e-05,
+      "loss": 0.3336,
+      "num_tokens": 6839414.0,
+      "step": 55
+    },
+    {
+      "epoch": 3.3404255319148937,
+      "grad_norm": 0.1529250136588493,
+      "learning_rate": 4.011417540268764e-05,
+      "loss": 0.3313,
+      "num_tokens": 7460094.0,
+      "step": 60
+    },
+    {
+      "epoch": 3.624113475177305,
+      "grad_norm": 0.13922308111530698,
+      "learning_rate": 3.8239287280718634e-05,
+      "loss": 0.3261,
+      "num_tokens": 8090603.0,
+      "step": 65
+    },
+    {
+      "epoch": 3.9078014184397163,
+      "grad_norm": 0.1510678073392735,
+      "learning_rate": 3.626225373056101e-05,
+      "loss": 0.3206,
+      "num_tokens": 8731464.0,
+      "step": 70
+    },
+    {
+      "epoch": 4.170212765957447,
+      "grad_norm": 0.13076324205354475,
+      "learning_rate": 3.420187906699333e-05,
+      "loss": 0.3138,
+      "num_tokens": 9337672.0,
+      "step": 75
+    },
+    {
+      "epoch": 4.453900709219858,
+      "grad_norm": 0.13075395705901488,
+      "learning_rate": 3.207776029368427e-05,
+      "loss": 0.3135,
+      "num_tokens": 9993032.0,
+      "step": 80
+    },
+    {
+      "epoch": 4.73758865248227,
+      "grad_norm": 0.1217000256711584,
+      "learning_rate": 2.99101007086695e-05,
+      "loss": 0.3126,
+      "num_tokens": 10628806.0,
+      "step": 85
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.17989296172670693,
+      "learning_rate": 2.7719517743133312e-05,
+      "loss": 0.3194,
+      "num_tokens": 11180570.0,
+      "step": 90
+    },
+    {
+      "epoch": 5.567375886524823,
+      "grad_norm": 0.12712404524211618,
+      "learning_rate": 2.5526846861212987e-05,
+      "loss": 0.3084,
+      "num_tokens": 618338.0,
+      "step": 95
+    },
+    {
+      "epoch": 5.851063829787234,
+      "grad_norm": 0.12153203094502081,
+      "learning_rate": 2.3352943386009223e-05,
+      "loss": 0.3051,
+      "num_tokens": 655360.0,
+      "step": 100
+    },
+    {
+      "epoch": 6.170212765957447,
+      "grad_norm": 0.1305667538126749,
+      "learning_rate": 2.1218484136710372e-05,
+      "loss": 0.3671,
+      "num_tokens": 1370427.0,
+      "step": 105
+    },
+    {
+      "epoch": 6.453900709219858,
+      "grad_norm": 0.11952808807930158,
+      "learning_rate": 1.914377076353511e-05,
+      "loss": 0.2998,
+      "num_tokens": 1995968.0,
+      "step": 110
+    },
+    {
+      "epoch": 6.73758865248227,
+      "grad_norm": 0.11957376304714733,
+      "learning_rate": 1.7148536651049078e-05,
+      "loss": 0.3056,
+      "num_tokens": 2640944.0,
+      "step": 115
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.1618112055166159,
+      "learning_rate": 1.5251759226471556e-05,
+      "loss": 0.3039,
+      "num_tokens": 3235538.0,
+      "step": 120
+    },
+    {
+      "epoch": 7.283687943262412,
+      "grad_norm": 0.12326180294670976,
+      "learning_rate": 1.3471479458178499e-05,
+      "loss": 0.2973,
+      "num_tokens": 3848771.0,
+      "step": 125
+    },
+    {
+      "epoch": 7.567375886524823,
+      "grad_norm": 0.12629369151907394,
+      "learning_rate": 1.1824630261220466e-05,
+      "loss": 0.301,
+      "num_tokens": 4504131.0,
+      "step": 130
+    },
+    {
+      "epoch": 7.851063829787234,
+      "grad_norm": 0.12537087846653183,
+      "learning_rate": 1.0326875441955191e-05,
+      "loss": 0.3028,
+      "num_tokens": 5139202.0,
+      "step": 135
+    },
+    {
+      "epoch": 8.113475177304965,
+      "grad_norm": 0.11821179377146723,
+      "learning_rate": 8.99246071365363e-06,
+      "loss": 0.3039,
+      "num_tokens": 5733796.0,
+      "step": 140
+    },
+    {
+      "epoch": 8.397163120567376,
+      "grad_norm": 0.12302203109401375,
+      "learning_rate": 7.834078200126287e-06,
+      "loss": 0.2976,
+      "num_tokens": 6355222.0,
+      "step": 145
+    },
+    {
+      "epoch": 8.680851063829786,
+      "grad_norm": 0.125688120347689,
+      "learning_rate": 6.862745716127312e-06,
+      "loss": 0.3,
+      "num_tokens": 6980870.0,
+      "step": 150
+    },
+    {
+      "epoch": 8.964539007092199,
+      "grad_norm": 0.12817483625488393,
+      "learning_rate": 6.087701972745999e-06,
+      "loss": 0.3018,
+      "num_tokens": 7625846.0,
+      "step": 155
+    },
+    {
+      "epoch": 9.22695035460993,
+      "grad_norm": 0.12206554634101204,
+      "learning_rate": 5.51631870452704e-06,
+      "loss": 0.2962,
+      "num_tokens": 8215782.0,
+      "step": 160
+    },
+    {
+      "epoch": 9.51063829787234,
+      "grad_norm": 0.11862499163677016,
+      "learning_rate": 5.154030554111651e-06,
+      "loss": 0.2983,
+      "num_tokens": 8860030.0,
+      "step": 165
+    },
+    {
+      "epoch": 9.794326241134751,
+      "grad_norm": 0.11740647423935624,
+      "learning_rate": 5.004283381293366e-06,
+      "loss": 0.2988,
+      "num_tokens": 9507891.0,
+      "step": 170
+    },
+    {
+      "epoch": 9.794326241134751,
+      "step": 170,
+      "total_flos": 385299641270272.0,
+      "train_loss": 0.13468429481281954,
+      "train_runtime": 769.9529,
+      "train_samples_per_second": 1.831,
+      "train_steps_per_second": 0.221
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 170,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 5,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 385299641270272.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}