Model save

README.md

@@ -0,0 +1,58 @@
+---
+base_model: Qwen/Qwen2.5-3B-Instruct
+library_name: transformers
+model_name: Qwen2.5-3B-Instruct-Distill-bs17k-fem32768-batch32-epoch3-8192
+tags:
+- generated_from_trainer
+- trl
+- sft
+licence: license
+---
+
+# Model Card for Qwen2.5-3B-Instruct-Distill-bs17k-fem32768-batch32-epoch3-8192
+
+This model is a fine-tuned version of [Qwen/Qwen2.5-3B-Instruct](https://huggingface.co/Qwen/Qwen2.5-3B-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="Lansechen/Qwen2.5-3B-Instruct-Distill-bs17k-fem32768-batch32-epoch3-8192", 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/chenran1995-the-chinese-university-of-hong-kong/huggingface/runs/qter43gp) 
+
+
+This model was trained with SFT.
+
+### Framework versions
+
+- TRL: 0.16.0.dev0
+- Transformers: 4.49.0
+- Pytorch: 2.5.1+cu121
+- Datasets: 3.3.1
+- 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 Gallouédec},
+	year         = 2020,
+	journal      = {GitHub repository},
+	publisher    = {GitHub},
+	howpublished = {\url{https://github.com/huggingface/trl}}
+}
+```

all_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 139688442593280.0,
+    "train_loss": 0.6959928004972397,
+    "train_runtime": 2479.5716,
+    "train_samples": 3794,
+    "train_samples_per_second": 4.873,
+    "train_steps_per_second": 0.038
+}

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": 1.05,
+  "temperature": 0.7,
+  "top_k": 20,
+  "top_p": 0.8,
+  "transformers_version": "4.49.0"
+}

train_results.json

@@ -0,0 +1,8 @@
+{
+    "total_flos": 139688442593280.0,
+    "train_loss": 0.6959928004972397,
+    "train_runtime": 2479.5716,
+    "train_samples": 3794,
+    "train_samples_per_second": 4.873,
+    "train_steps_per_second": 0.038
+}

trainer_state.json

@@ -0,0 +1,187 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9206349206349205,
+  "eval_steps": 500,
+  "global_step": 93,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.15873015873015872,
+      "grad_norm": 3.1405258178710938,
+      "learning_rate": 5e-05,
+      "loss": 0.9988,
+      "mean_token_accuracy": 0.7276985734701157,
+      "step": 5
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 0.8156048059463501,
+      "learning_rate": 4.964250209756261e-05,
+      "loss": 0.8915,
+      "mean_token_accuracy": 0.7418786332011222,
+      "step": 10
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.5297970175743103,
+      "learning_rate": 4.8581368812494645e-05,
+      "loss": 0.8028,
+      "mean_token_accuracy": 0.759541466832161,
+      "step": 15
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 0.4519253075122833,
+      "learning_rate": 4.6850320404673266e-05,
+      "loss": 0.7684,
+      "mean_token_accuracy": 0.7664183840155602,
+      "step": 20
+    },
+    {
+      "epoch": 0.7936507936507936,
+      "grad_norm": 0.38354581594467163,
+      "learning_rate": 4.450436542297082e-05,
+      "loss": 0.747,
+      "mean_token_accuracy": 0.7712402656674385,
+      "step": 25
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.32026010751724243,
+      "learning_rate": 4.161805266606917e-05,
+      "loss": 0.7254,
+      "mean_token_accuracy": 0.7770746260881424,
+      "step": 30
+    },
+    {
+      "epoch": 1.0952380952380953,
+      "grad_norm": 0.38725751638412476,
+      "learning_rate": 3.8283102201201284e-05,
+      "loss": 0.6948,
+      "mean_token_accuracy": 0.7841671456893285,
+      "step": 35
+    },
+    {
+      "epoch": 1.253968253968254,
+      "grad_norm": 0.33806172013282776,
+      "learning_rate": 3.460549072133806e-05,
+      "loss": 0.6656,
+      "mean_token_accuracy": 0.7906794264912606,
+      "step": 40
+    },
+    {
+      "epoch": 1.4126984126984126,
+      "grad_norm": 0.311052531003952,
+      "learning_rate": 3.070208386114892e-05,
+      "loss": 0.6712,
+      "mean_token_accuracy": 0.7883317843079567,
+      "step": 45
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.2916868031024933,
+      "learning_rate": 2.6696922488622945e-05,
+      "loss": 0.6636,
+      "mean_token_accuracy": 0.7903473630547524,
+      "step": 50
+    },
+    {
+      "epoch": 1.7301587301587302,
+      "grad_norm": 0.2753700315952301,
+      "learning_rate": 2.2717280985058025e-05,
+      "loss": 0.6436,
+      "mean_token_accuracy": 0.7960700690746307,
+      "step": 55
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.2455938756465912,
+      "learning_rate": 1.888962277178548e-05,
+      "loss": 0.6392,
+      "mean_token_accuracy": 0.7972360864281655,
+      "step": 60
+    },
+    {
+      "epoch": 2.0317460317460316,
+      "grad_norm": 0.3443191349506378,
+      "learning_rate": 1.5335581607249064e-05,
+      "loss": 0.6408,
+      "mean_token_accuracy": 0.7964212745428085,
+      "step": 65
+    },
+    {
+      "epoch": 2.1904761904761907,
+      "grad_norm": 0.2579975724220276,
+      "learning_rate": 1.216809635913745e-05,
+      "loss": 0.6068,
+      "mean_token_accuracy": 0.8054686427116394,
+      "step": 70
+    },
+    {
+      "epoch": 2.3492063492063493,
+      "grad_norm": 0.26707184314727783,
+      "learning_rate": 9.487822079201895e-06,
+      "loss": 0.6046,
+      "mean_token_accuracy": 0.8058722853660584,
+      "step": 75
+    },
+    {
+      "epoch": 2.507936507936508,
+      "grad_norm": 0.24455317854881287,
+      "learning_rate": 7.379931428156497e-06,
+      "loss": 0.6065,
+      "mean_token_accuracy": 0.8047785997390747,
+      "step": 80
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.2308081090450287,
+      "learning_rate": 5.911408093673812e-06,
+      "loss": 0.6034,
+      "mean_token_accuracy": 0.8056193083524704,
+      "step": 85
+    },
+    {
+      "epoch": 2.825396825396825,
+      "grad_norm": 0.22246606647968292,
+      "learning_rate": 5.12891821013231e-06,
+      "loss": 0.6008,
+      "mean_token_accuracy": 0.8066127508878708,
+      "step": 90
+    },
+    {
+      "epoch": 2.9206349206349205,
+      "mean_token_accuracy": 0.8056537633140882,
+      "step": 93,
+      "total_flos": 139688442593280.0,
+      "train_loss": 0.6959928004972397,
+      "train_runtime": 2479.5716,
+      "train_samples_per_second": 4.873,
+      "train_steps_per_second": 0.038
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 93,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "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": 139688442593280.0,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}