Model save

README.md

@@ -0,0 +1,68 @@
+---
+base_model: Qwen/Qwen2.5-1.5B-Instruct
+library_name: transformers
+model_name: Qwen-2.5-1.5B-Instruct-Datamix
+tags:
+- generated_from_trainer
+- trl
+- grpo
+licence: license
+---
+
+# Model Card for Qwen-2.5-1.5B-Instruct-Datamix
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-1.5B-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="andrewsiah/Qwen-2.5-1.5B-Instruct-Datamix", 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/hong-exploration/openr1/runs/x1urdcd4) 
+
+
+This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
+
+### Framework versions
+
+- TRL: 0.15.0.dev0
+- Transformers: 4.49.0.dev0
+- Pytorch: 2.5.1
+- Datasets: 3.2.0
+- Tokenizers: 0.21.0
+
+## Citations
+
+Cite GRPO as:
+
+```bibtex
+@article{zhihong2024deepseekmath,
+    title        = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
+    author       = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
+    year         = 2024,
+    eprint       = {arXiv:2402.03300},
+}
+
+```
+
+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 Gallouédec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

all_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 0.0,
+    "train_loss": 1.0761486015149526,
+    "train_runtime": 1434.8598,
+    "train_samples": 2000,
+    "train_samples_per_second": 1.394,
+    "train_steps_per_second": 0.005
+}

generation_config.json

@@ -0,0 +1,14 @@
+{
+  "bos_token_id": 151643,
+  "do_sample": true,
+  "eos_token_id": [
+    151645,
+    151643
+  ],
+  "pad_token_id": 151643,
+  "repetition_penalty": 1.1,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.49.0.dev0"
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 0.0,
+    "train_loss": 1.0761486015149526,
+    "train_runtime": 1434.8598,
+    "train_samples": 2000,
+    "train_samples_per_second": 1.394,
+    "train_steps_per_second": 0.005
+}

trainer_state.json

@@ -0,0 +1,89 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.8888888888888888,
+  "eval_steps": 4,
+  "global_step": 7,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval_completion_length": 362.30074795809657,
+      "eval_kl": 0.0,
+      "eval_loss": -9.876328113023192e-05,
+      "eval_reward": 0.665381520986557,
+      "eval_reward_std": 0.4824826947667382,
+      "eval_rewards/accuracy_reward": 0.13372565534981815,
+      "eval_rewards/format_reward": 0.5316558669913899,
+      "eval_runtime": 333.2158,
+      "eval_samples_per_second": 2.101,
+      "eval_steps_per_second": 0.012,
+      "step": 0
+    },
+    {
+      "epoch": 0.5079365079365079,
+      "eval_completion_length": 217.30788110803675,
+      "eval_kl": 28.520580150462962,
+      "eval_loss": 0.09147126972675323,
+      "eval_reward": 0.923280468693486,
+      "eval_reward_std": 0.3771378709762185,
+      "eval_rewards/accuracy_reward": 0.13583003308762004,
+      "eval_rewards/format_reward": 0.7874504283622459,
+      "eval_runtime": 251.0843,
+      "eval_samples_per_second": 2.788,
+      "eval_steps_per_second": 0.016,
+      "step": 4
+    },
+    {
+      "completion_length": 208.86775398254395,
+      "epoch": 0.6349206349206349,
+      "grad_norm": 73.1821060180664,
+      "kl": 1.2861328125,
+      "learning_rate": 5.000000000000003e-06,
+      "loss": 1.4987,
+      "reward": 0.9737723916769028,
+      "reward_std": 0.3762528672814369,
+      "rewards/accuracy_reward": 0.12611607741564512,
+      "rewards/format_reward": 0.8476562947034836,
+      "step": 5
+    },
+    {
+      "completion_length": 188.7966046333313,
+      "epoch": 0.8888888888888888,
+      "kl": 0.51080322265625,
+      "reward": 0.9969308339059353,
+      "reward_std": 0.3559337202459574,
+      "rewards/accuracy_reward": 0.1283482201397419,
+      "rewards/format_reward": 0.8685826286673546,
+      "step": 7,
+      "total_flos": 0.0,
+      "train_loss": 1.0761486015149526,
+      "train_runtime": 1434.8598,
+      "train_samples_per_second": 1.394,
+      "train_steps_per_second": 0.005
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 7,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2,
+  "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": 32,
+  "trial_name": null,
+  "trial_params": null
+}