Model save

README.md

@@ -1,11 +1,9 @@
 ---
 base_model: Qwen/Qwen2.5-1.5B-Instruct
-datasets: HuggingFaceH4/Bespoke-Stratos-17k
 library_name: transformers
 model_name: Qwen2.5-1.5B-Open-R1-Distill
 tags:
 - generated_from_trainer
-- open-r1
 - trl
 - sft
 licence: license
@@ -13,7 +11,7 @@ licence: license
 
 # Model Card for Qwen2.5-1.5B-Open-R1-Distill
 
-This model is a fine-tuned version of [Qwen/Qwen2.5-1.5B-Instruct](https://huggingface.co/Qwen/Qwen2.5-1.5B-Instruct) on the [HuggingFaceH4/Bespoke-Stratos-17k](https://huggingface.co/datasets/HuggingFaceH4/Bespoke-Stratos-17k) dataset.
+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
@@ -29,14 +27,14 @@ 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/xiaojia-song12/huggingface/runs/hc5miacf) 
+[<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/xiaojia-song12/huggingface/runs/977ijn5q) 
 
 
 This model was trained with SFT.
 
 ### Framework versions
 
-- TRL: 0.15.0.dev0
+- TRL: 0.16.0.dev0
 - Transformers: 4.49.0.dev0
 - Pytorch: 2.7.0a0+git3a58512
 - Datasets: 3.2.0

all_results.json

@@ -1,13 +1,8 @@
 {
-    "eval_loss": 0.8572873473167419,
-    "eval_runtime": 2.3373,
-    "eval_samples": 100,
-    "eval_samples_per_second": 55.192,
-    "eval_steps_per_second": 2.139,
     "total_flos": 6.966137809639834e+17,
-    "train_loss": 0.8674715298872727,
-    "train_runtime": 815.3952,
+    "train_loss": 0.8673680914929632,
+    "train_runtime": 795.9713,
     "train_samples": 16610,
-    "train_samples_per_second": 26.518,
-    "train_steps_per_second": 0.207
+    "train_samples_per_second": 27.166,
+    "train_steps_per_second": 0.212
 }

config.json

@@ -23,7 +23,7 @@
   "tie_word_embeddings": true,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.49.0.dev0",
-  "use_cache": true,
+  "use_cache": false,
   "use_sliding_window": false,
   "vocab_size": 151936
 }

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d69e13d1e0a1431e99e002eb37cd9f5b7404337bcbd336a59d7f02b2e4bbadf7
+oid sha256:43f380d7be9586da69712b78466b8c4247dc743323c393ed4e447e8943bead8b
 size 3087467144

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 6.966137809639834e+17,
-    "train_loss": 0.8674715298872727,
-    "train_runtime": 815.3952,
+    "train_loss": 0.8673680914929632,
+    "train_runtime": 795.9713,
     "train_samples": 16610,
-    "train_samples_per_second": 26.518,
-    "train_steps_per_second": 0.207
+    "train_samples_per_second": 27.166,
+    "train_steps_per_second": 0.212
 }

trainer_state.json

@@ -10,286 +10,286 @@
   "log_history": [
     {
       "epoch": 0.029585798816568046,
-      "grad_norm": 5.03125,
+      "grad_norm": 2.40625,
       "learning_rate": 5.882352941176471e-06,
       "loss": 1.1011,
-      "mean_token_accuracy": 0.7099399748962237,
+      "mean_token_accuracy": 0.7099346258467769,
       "step": 5
     },
     {
       "epoch": 0.05917159763313609,
-      "grad_norm": 1.484375,
+      "grad_norm": 1.46875,
       "learning_rate": 1.1764705882352942e-05,
-      "loss": 1.0741,
-      "mean_token_accuracy": 0.7142575641587607,
+      "loss": 1.0735,
+      "mean_token_accuracy": 0.7143598902988348,
       "step": 10
     },
     {
       "epoch": 0.08875739644970414,
-      "grad_norm": 0.9296875,
+      "grad_norm": 0.93359375,
       "learning_rate": 1.7647058823529414e-05,
-      "loss": 1.025,
-      "mean_token_accuracy": 0.7212255669151615,
+      "loss": 1.0246,
+      "mean_token_accuracy": 0.7213492548758127,
       "step": 15
     },
     {
       "epoch": 0.11834319526627218,
       "grad_norm": 0.61328125,
       "learning_rate": 1.9980782984658682e-05,
-      "loss": 0.9455,
-      "mean_token_accuracy": 0.7365584053166825,
+      "loss": 0.9456,
+      "mean_token_accuracy": 0.7365624101438417,
       "step": 20
     },
     {
       "epoch": 0.14792899408284024,
-      "grad_norm": 0.515625,
+      "grad_norm": 0.51171875,
       "learning_rate": 1.9863613034027224e-05,
-      "loss": 0.9097,
-      "mean_token_accuracy": 0.7419796471746476,
+      "loss": 0.9101,
+      "mean_token_accuracy": 0.7418235078195066,
       "step": 25
     },
     {
       "epoch": 0.17751479289940827,
-      "grad_norm": 0.42578125,
+      "grad_norm": 0.423828125,
       "learning_rate": 1.9641197940012136e-05,
-      "loss": 0.8953,
-      "mean_token_accuracy": 0.744632902913685,
+      "loss": 0.8958,
+      "mean_token_accuracy": 0.7445377548880543,
       "step": 30
     },
     {
       "epoch": 0.20710059171597633,
       "grad_norm": 0.408203125,
       "learning_rate": 1.9315910880512792e-05,
-      "loss": 0.8769,
-      "mean_token_accuracy": 0.7474371048323014,
+      "loss": 0.8774,
+      "mean_token_accuracy": 0.7473939577517185,
       "step": 35
     },
     {
       "epoch": 0.23668639053254437,
-      "grad_norm": 0.388671875,
+      "grad_norm": 0.390625,
       "learning_rate": 1.8891222681391853e-05,
-      "loss": 0.8623,
-      "mean_token_accuracy": 0.7511180018113048,
+      "loss": 0.8628,
+      "mean_token_accuracy": 0.7510309574773866,
       "step": 40
     },
     {
       "epoch": 0.26627218934911245,
-      "grad_norm": 0.357421875,
+      "grad_norm": 4.75,
       "learning_rate": 1.8371664782625287e-05,
-      "loss": 0.8557,
-      "mean_token_accuracy": 0.7520890468920156,
+      "loss": 0.8562,
+      "mean_token_accuracy": 0.7519657291179588,
       "step": 45
     },
     {
       "epoch": 0.2958579881656805,
-      "grad_norm": 0.326171875,
+      "grad_norm": 0.322265625,
       "learning_rate": 1.7762780887657576e-05,
-      "loss": 0.8519,
-      "mean_token_accuracy": 0.7527647439023715,
+      "loss": 0.8526,
+      "mean_token_accuracy": 0.7526042764494737,
       "step": 50
     },
     {
       "epoch": 0.3254437869822485,
-      "grad_norm": 0.34375,
+      "grad_norm": 0.33984375,
       "learning_rate": 1.7071067811865477e-05,
-      "loss": 0.8499,
-      "mean_token_accuracy": 0.7530196620772068,
+      "loss": 0.8506,
+      "mean_token_accuracy": 0.7529147428713211,
       "step": 55
     },
     {
       "epoch": 0.35502958579881655,
-      "grad_norm": 0.345703125,
+      "grad_norm": 0.349609375,
       "learning_rate": 1.6303906161279554e-05,
-      "loss": 0.8623,
-      "mean_token_accuracy": 0.7496151012147616,
+      "loss": 0.8627,
+      "mean_token_accuracy": 0.7494820701814584,
       "step": 60
     },
     {
       "epoch": 0.38461538461538464,
-      "grad_norm": 0.322265625,
+      "grad_norm": 0.32421875,
       "learning_rate": 1.5469481581224274e-05,
-      "loss": 0.8491,
-      "mean_token_accuracy": 0.7524191922540802,
+      "loss": 0.8496,
+      "mean_token_accuracy": 0.7523080532405824,
       "step": 65
     },
     {
       "epoch": 0.41420118343195267,
-      "grad_norm": 0.33984375,
+      "grad_norm": 0.3359375,
       "learning_rate": 1.4576697415156818e-05,
-      "loss": 0.8418,
-      "mean_token_accuracy": 0.7547066041239219,
+      "loss": 0.8422,
+      "mean_token_accuracy": 0.7546321496781302,
       "step": 70
     },
     {
       "epoch": 0.4437869822485207,
       "grad_norm": 0.3125,
       "learning_rate": 1.3635079705638298e-05,
-      "loss": 0.8416,
-      "mean_token_accuracy": 0.7544158237788706,
+      "loss": 0.842,
+      "mean_token_accuracy": 0.7544247004239197,
       "step": 75
     },
     {
       "epoch": 0.47337278106508873,
-      "grad_norm": 9.5625,
+      "grad_norm": 0.322265625,
       "learning_rate": 1.2654675551080724e-05,
-      "loss": 0.8451,
-      "mean_token_accuracy": 0.7531590392914497,
+      "loss": 0.8453,
+      "mean_token_accuracy": 0.7530717828212506,
       "step": 80
     },
     {
       "epoch": 0.5029585798816568,
-      "grad_norm": 0.328125,
+      "grad_norm": 0.330078125,
       "learning_rate": 1.164594590280734e-05,
       "loss": 0.8455,
-      "mean_token_accuracy": 0.7532937761410722,
+      "mean_token_accuracy": 0.7532910964218957,
       "step": 85
     },
     {
       "epoch": 0.5325443786982249,
-      "grad_norm": 0.337890625,
+      "grad_norm": 0.333984375,
       "learning_rate": 1.0619653946285948e-05,
-      "loss": 0.8399,
-      "mean_token_accuracy": 0.7544911735347861,
+      "loss": 0.8396,
+      "mean_token_accuracy": 0.7545698270600318,
       "step": 90
     },
     {
       "epoch": 0.5621301775147929,
-      "grad_norm": 0.35546875,
+      "grad_norm": 0.353515625,
       "learning_rate": 9.586750257511868e-06,
-      "loss": 0.8358,
-      "mean_token_accuracy": 0.7556337403583416,
+      "loss": 0.8353,
+      "mean_token_accuracy": 0.7558067680588229,
       "step": 95
     },
     {
       "epoch": 0.591715976331361,
-      "grad_norm": 13.75,
+      "grad_norm": 0.32421875,
       "learning_rate": 8.558255959926533e-06,
-      "loss": 0.8269,
-      "mean_token_accuracy": 0.7575153177712692,
+      "loss": 0.8265,
+      "mean_token_accuracy": 0.7575928127928315,
       "step": 100
     },
     {
       "epoch": 0.591715976331361,
-      "eval_loss": 0.8590129017829895,
-      "eval_mean_token_accuracy": 0.7343020822802343,
-      "eval_runtime": 2.301,
-      "eval_samples_per_second": 56.063,
-      "eval_steps_per_second": 2.173,
+      "eval_loss": 0.858591616153717,
+      "eval_mean_token_accuracy": 0.7338160377061327,
+      "eval_runtime": 2.2821,
+      "eval_samples_per_second": 56.526,
+      "eval_steps_per_second": 2.191,
       "step": 100
     },
     {
       "epoch": 0.621301775147929,
-      "grad_norm": 0.326171875,
+      "grad_norm": 0.3203125,
       "learning_rate": 7.545145128592009e-06,
-      "loss": 0.8289,
-      "mean_token_accuracy": 0.7575234407641943,
+      "loss": 0.8284,
+      "mean_token_accuracy": 0.7576165643011782,
       "step": 105
     },
     {
       "epoch": 0.650887573964497,
-      "grad_norm": 21.25,
+      "grad_norm": 0.314453125,
       "learning_rate": 6.558227696373617e-06,
-      "loss": 0.8501,
-      "mean_token_accuracy": 0.7518221200713989,
+      "loss": 0.8497,
+      "mean_token_accuracy": 0.7519416205438242,
       "step": 110
     },
     {
       "epoch": 0.6804733727810651,
       "grad_norm": 0.322265625,
       "learning_rate": 5.608034111526298e-06,
-      "loss": 0.8361,
-      "mean_token_accuracy": 0.7556435102369359,
+      "loss": 0.8356,
+      "mean_token_accuracy": 0.7557316956442416,
       "step": 115
     },
     {
       "epoch": 0.7100591715976331,
       "grad_norm": 0.328125,
       "learning_rate": 4.704702977392914e-06,
-      "loss": 0.8234,
-      "mean_token_accuracy": 0.7592704800654426,
+      "loss": 0.8229,
+      "mean_token_accuracy": 0.7593895998895455,
       "step": 120
     },
     {
       "epoch": 0.7396449704142012,
       "grad_norm": 0.314453125,
       "learning_rate": 3.857872873103322e-06,
-      "loss": 0.8328,
-      "mean_token_accuracy": 0.7565933236472202,
+      "loss": 0.8322,
+      "mean_token_accuracy": 0.756736495558991,
       "step": 125
     },
     {
       "epoch": 0.7692307692307693,
-      "grad_norm": 0.328125,
+      "grad_norm": 0.326171875,
       "learning_rate": 3.0765795095517026e-06,
-      "loss": 0.8372,
-      "mean_token_accuracy": 0.7552126694771518,
+      "loss": 0.8367,
+      "mean_token_accuracy": 0.7552458819622366,
       "step": 130
     },
     {
       "epoch": 0.7988165680473372,
-      "grad_norm": 0.314453125,
+      "grad_norm": 0.3125,
       "learning_rate": 2.369159318001937e-06,
-      "loss": 0.8437,
-      "mean_token_accuracy": 0.7534241447974153,
+      "loss": 0.8431,
+      "mean_token_accuracy": 0.753608536029534,
       "step": 135
     },
     {
       "epoch": 0.8284023668639053,
       "grad_norm": 0.318359375,
       "learning_rate": 1.743160500034443e-06,
-      "loss": 0.835,
-      "mean_token_accuracy": 0.7563463400384239,
+      "loss": 0.8345,
+      "mean_token_accuracy": 0.7563751731445558,
       "step": 140
     },
     {
       "epoch": 0.8579881656804734,
       "grad_norm": 0.330078125,
       "learning_rate": 1.2052624879351105e-06,
-      "loss": 0.8351,
-      "mean_token_accuracy": 0.7560251329699866,
+      "loss": 0.8346,
+      "mean_token_accuracy": 0.7562260551940483,
       "step": 145
     },
     {
       "epoch": 0.8875739644970414,
       "grad_norm": 0.30859375,
       "learning_rate": 7.612046748871327e-07,
-      "loss": 0.8254,
-      "mean_token_accuracy": 0.7586063543068307,
+      "loss": 0.8249,
+      "mean_token_accuracy": 0.7586945500381791,
       "step": 150
     },
     {
       "epoch": 0.9171597633136095,
       "grad_norm": 0.31640625,
       "learning_rate": 4.1572517541747294e-07,
-      "loss": 0.8158,
-      "mean_token_accuracy": 0.7611673621269803,
+      "loss": 0.8153,
+      "mean_token_accuracy": 0.7613143505011888,
       "step": 155
     },
     {
       "epoch": 0.9467455621301775,
-      "grad_norm": 0.32421875,
+      "grad_norm": 0.322265625,
       "learning_rate": 1.7251026952640583e-07,
-      "loss": 0.8289,
-      "mean_token_accuracy": 0.7576611812663976,
+      "loss": 0.8284,
+      "mean_token_accuracy": 0.7578062563797692,
       "step": 160
     },
     {
       "epoch": 0.9763313609467456,
       "grad_norm": 0.3125,
       "learning_rate": 3.4155069933301535e-08,
-      "loss": 0.8216,
-      "mean_token_accuracy": 0.7594309628290168,
+      "loss": 0.8211,
+      "mean_token_accuracy": 0.759621495589895,
       "step": 165
     },
     {
       "epoch": 1.0,
-      "mean_token_accuracy": 0.7546930193447599,
+      "mean_token_accuracy": 0.7546903746827489,
       "step": 169,
       "total_flos": 6.966137809639834e+17,
-      "train_loss": 0.8674715298872727,
-      "train_runtime": 815.3952,
-      "train_samples_per_second": 26.518,
-      "train_steps_per_second": 0.207
+      "train_loss": 0.8673680914929632,
+      "train_runtime": 795.9713,
+      "train_samples_per_second": 27.166,
+      "train_steps_per_second": 0.212
     }
   ],
   "logging_steps": 5,

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1fd4518eeebb19a79bf77cb8690174a32d07f567e7bd1ae9d9a31d0e6e6e1074
+oid sha256:fbd31b7f95fae03b8da98d830088e494be82c7e417e3e2998543438d39a57832
 size 6008