Model save

README.md

@@ -0,0 +1,69 @@
+---
+base_model: Rudra-ai/Llama-3.2-3B-Instruct-cot-19-11-1
+library_name: transformers
+model_name: Llama-3.2-3B-Instruct-cot-19-11-2
+tags:
+- generated_from_trainer
+- trl
+- dpo
+- unsloth
+licence: license
+---
+
+# Model Card for Llama-3.2-3B-Instruct-cot-19-11-2
+
+This model is a fine-tuned version of [Rudra-ai/Llama-3.2-3B-Instruct-cot-19-11-1](https://huggingface.co/Rudra-ai/Llama-3.2-3B-Instruct-cot-19-11-1).
+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="Rudra-ai/Llama-3.2-3B-Instruct-cot-19-11-2", device="cuda")
+output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
+print(output["generated_text"])
+```
+
+## Training procedure
+
+
+
+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: 0.12.1
+- Transformers: 4.46.1
+- Pytorch: 2.4.1+cu124
+- Datasets: 3.1.0
+- Tokenizers: 0.20.3
+
+## 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 Gallouédec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

all_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 1.0,
+    "total_flos": 0.0,
+    "train_loss": 2.0509657764434817,
+    "train_runtime": 2877.0971,
+    "train_samples": 2000,
+    "train_samples_per_second": 0.695,
+    "train_steps_per_second": 0.017
+}

generation_config.json

@@ -0,0 +1,8 @@
+{
+  "bos_token_id": 128000,
+  "do_sample": true,
+  "max_length": 131072,
+  "temperature": 0.6,
+  "top_p": 0.9,
+  "transformers_version": "4.46.1"
+}

train_results.json

@@ -0,0 +1,9 @@
+{
+    "epoch": 1.0,
+    "total_flos": 0.0,
+    "train_loss": 2.0509657764434817,
+    "train_runtime": 2877.0971,
+    "train_samples": 2000,
+    "train_samples_per_second": 0.695,
+    "train_steps_per_second": 0.017
+}

trainer_state.json

@@ -0,0 +1,132 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 50,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.02,
+      "grad_norm": 2673.435546875,
+      "learning_rate": 2.0000000000000003e-06,
+      "logits/chosen": 0.6171875,
+      "logits/rejected": 0.294921875,
+      "logps/chosen": -414.0,
+      "logps/rejected": -478.0,
+      "loss": 6.9141,
+      "rewards/accuracies": 0.0,
+      "rewards/chosen": 0.0,
+      "rewards/margins": 0.0,
+      "rewards/rejected": 0.0,
+      "step": 1
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 368.8966369628906,
+      "learning_rate": 9.698463103929542e-06,
+      "logits/chosen": 0.58203125,
+      "logits/rejected": 0.41015625,
+      "logps/chosen": -370.0,
+      "logps/rejected": -496.0,
+      "loss": 3.4334,
+      "rewards/accuracies": 0.7666666507720947,
+      "rewards/chosen": -3.03125,
+      "rewards/margins": 6.5,
+      "rewards/rejected": -9.5625,
+      "step": 10
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 122.28874969482422,
+      "learning_rate": 7.500000000000001e-06,
+      "logits/chosen": 0.8984375,
+      "logits/rejected": 0.75390625,
+      "logps/chosen": -560.0,
+      "logps/rejected": -832.0,
+      "loss": 2.7651,
+      "rewards/accuracies": 0.9700000286102295,
+      "rewards/chosen": -20.25,
+      "rewards/margins": 23.25,
+      "rewards/rejected": -43.5,
+      "step": 20
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 270.5760498046875,
+      "learning_rate": 4.131759111665349e-06,
+      "logits/chosen": 1.1015625,
+      "logits/rejected": 0.93359375,
+      "logps/chosen": -560.0,
+      "logps/rejected": -808.0,
+      "loss": 1.6426,
+      "rewards/accuracies": 1.0,
+      "rewards/chosen": -19.75,
+      "rewards/margins": 21.25,
+      "rewards/rejected": -41.0,
+      "step": 30
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 198.14657592773438,
+      "learning_rate": 1.1697777844051105e-06,
+      "logits/chosen": 1.109375,
+      "logits/rejected": 0.91015625,
+      "logps/chosen": -500.0,
+      "logps/rejected": -780.0,
+      "loss": 0.3845,
+      "rewards/accuracies": 0.9800000190734863,
+      "rewards/chosen": -16.625,
+      "rewards/margins": 20.875,
+      "rewards/rejected": -37.5,
+      "step": 40
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 122.92610931396484,
+      "learning_rate": 0.0,
+      "logits/chosen": 1.0625,
+      "logits/rejected": 0.75,
+      "logps/chosen": -520.0,
+      "logps/rejected": -836.0,
+      "loss": 1.6812,
+      "rewards/accuracies": 0.9800000190734863,
+      "rewards/chosen": -17.125,
+      "rewards/margins": 24.5,
+      "rewards/rejected": -41.75,
+      "step": 50
+    },
+    {
+      "epoch": 1.0,
+      "step": 50,
+      "total_flos": 0.0,
+      "train_loss": 2.0509657764434817,
+      "train_runtime": 2877.0971,
+      "train_samples_per_second": 0.695,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 50,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "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": 1,
+  "trial_name": null,
+  "trial_params": null
+}