Upload checkpoint 260

README.md

@@ -2,7 +2,7 @@
 base_model: InfiniAILab/OpenR1-Qwen-3B-SFT-Instruct
 library_name: peft
 ---
-# T2 3B Instruct (Step 250 Checkpoint)
+# T2 3B Instruct (Step 260 Checkpoint)
 
 > [!NOTE]
 > Training in progress...
@@ -15,11 +15,11 @@ library_name: peft
 </head>
 <body>
 <div style="width: 100%; background-color: #e0e0e0; border-radius: 25px; overflow: hidden; margin: 20px 0;">
-  <div style="height: 30px; width: 0.97%; background-color: #44965a; text-align: center; line-height: 30px; color: white; border-radius: 25px 0 0 25px;">
+  <div style="height: 30px; width: 1.01%; background-color: #44965a; text-align: center; line-height: 30px; color: white; border-radius: 25px 0 0 25px;">
     <!-- 1.0% -->
   </div>
 </div>
-<p style="font-family: Arial, sans-serif; font-size: 16px;">Progress: 250 out of 25777 steps</p>
+<p style="font-family: Arial, sans-serif; font-size: 16px;">Progress: 260 out of 25777 steps</p>
 </body>
 </html>
 

adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3b147f7999557344770a2cd73c395986545692fdec8c0ea13a8289e03cee5570
+oid sha256:f36d4317630c5110dbff946f802587ea3bc701a60001465c6febcf3fcacf9375
 size 479005064

optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8f4a6893aabf45795b45777fe1c6f2f0f453da25f72335dd46750f1abdc76e7f
-size 245114274
+oid sha256:b9848c53fab9a1e00c00ba4b2d7d20e2a0495d5aee7f8511fbec43cb6e7e0efc
+size 245114786

rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3484f92b33a888f5c310ea0cec056ab72ee15bba0e7fd5601a9c96c68e14b094
+oid sha256:5278935d54b157c63f6fe795ae983618d6916455dc631a48cf005af6f372a998
 size 14244

scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:35923cae999208ef8aef781ab64712bb4ecd3626f9fd5c318d5708ec862833ef
+oid sha256:0be909e7c2c0b47ba64df5a06fb5d207fdb83522276bdf706943c41b96e047c6
 size 1064

trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.00969838037047813,
+  "epoch": 0.010086315585297255,
   "eval_steps": 500,
-  "global_step": 250,
+  "global_step": 260,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -3758,6 +3758,156 @@
       "rewards/strict_format_reward_func": 0.028125000884756446,
       "rewards/thinkcount_reward_func": 0.036503125447779894,
       "step": 250
+    },
+    {
+      "completion_length": 8101.75,
+      "epoch": 0.009737173891960042,
+      "grad_norm": 2.8747098445892334,
+      "kl": 0.010620866902172565,
+      "learning_rate": 4.848719937936385e-07,
+      "loss": 0.0004,
+      "reward": 0.849343677982688,
+      "reward_std": 0.8998125595971942,
+      "rewards/correctness_reward_func": 0.76171875,
+      "rewards/soft_format_reward_func": 0.018750000279396772,
+      "rewards/strict_format_reward_func": 0.034375000512227416,
+      "rewards/thinkcount_reward_func": 0.034500000067055225,
+      "step": 251
+    },
+    {
+      "completion_length": 8098.25,
+      "epoch": 0.009775967413441956,
+      "grad_norm": 7.176968097686768,
+      "kl": 0.01608081115409732,
+      "learning_rate": 4.86811481768813e-07,
+      "loss": 0.0006,
+      "reward": 0.36659060791134834,
+      "reward_std": 0.517820986919105,
+      "rewards/correctness_reward_func": 0.275390625,
+      "rewards/soft_format_reward_func": 0.026562500395812094,
+      "rewards/strict_format_reward_func": 0.02812500041909516,
+      "rewards/thinkcount_reward_func": 0.03651250060647726,
+      "step": 252
+    },
+    {
+      "completion_length": 7920.53125,
+      "epoch": 0.009814760934923868,
+      "grad_norm": 10.816934585571289,
+      "kl": 0.03997403662651777,
+      "learning_rate": 4.887509697439876e-07,
+      "loss": 0.0016,
+      "reward": 0.6573624610900879,
+      "reward_std": 0.5487498492002487,
+      "rewards/correctness_reward_func": 0.544921875,
+      "rewards/soft_format_reward_func": 0.03437500074505806,
+      "rewards/strict_format_reward_func": 0.053125000558793545,
+      "rewards/thinkcount_reward_func": 0.024940625298768282,
+      "step": 253
+    },
+    {
+      "completion_length": 8112.5,
+      "epoch": 0.00985355445640578,
+      "grad_norm": 4.7223100662231445,
+      "kl": 0.015217800391837955,
+      "learning_rate": 4.906904577191622e-07,
+      "loss": 0.0006,
+      "reward": 0.6937499912455678,
+      "reward_std": 0.5481314389035106,
+      "rewards/correctness_reward_func": 0.609375,
+      "rewards/soft_format_reward_func": 0.018750000395812094,
+      "rewards/strict_format_reward_func": 0.031250000232830644,
+      "rewards/thinkcount_reward_func": 0.03437500027939677,
+      "step": 254
+    },
+    {
+      "completion_length": 7902.46875,
+      "epoch": 0.009892347977887692,
+      "grad_norm": 2.7174127101898193,
+      "kl": 0.010821214877068996,
+      "learning_rate": 4.926299456943368e-07,
+      "loss": 0.0004,
+      "reward": 0.7515398059040308,
+      "reward_std": 0.48249518871307373,
+      "rewards/correctness_reward_func": 0.66162109375,
+      "rewards/soft_format_reward_func": 0.020312500186264515,
+      "rewards/strict_format_reward_func": 0.03437500074505806,
+      "rewards/thinkcount_reward_func": 0.03523125033825636,
+      "step": 255
+    },
+    {
+      "completion_length": 8106.75,
+      "epoch": 0.009931141499369606,
+      "grad_norm": 8.932234764099121,
+      "kl": 0.05964707024395466,
+      "learning_rate": 4.945694336695113e-07,
+      "loss": 0.0024,
+      "reward": 0.626660130918026,
+      "reward_std": 0.6535767344757915,
+      "rewards/correctness_reward_func": 0.53759765625,
+      "rewards/soft_format_reward_func": 0.02187500020954758,
+      "rewards/strict_format_reward_func": 0.03437500004656613,
+      "rewards/thinkcount_reward_func": 0.03281250037252903,
+      "step": 256
+    },
+    {
+      "completion_length": 8076.5,
+      "epoch": 0.009969935020851518,
+      "grad_norm": 6.399128437042236,
+      "kl": 0.01735225645825267,
+      "learning_rate": 4.965089216446859e-07,
+      "loss": 0.0007,
+      "reward": 0.36347654554992914,
+      "reward_std": 0.5175247741863132,
+      "rewards/correctness_reward_func": 0.2822265625,
+      "rewards/soft_format_reward_func": 0.020312500768341124,
+      "rewards/strict_format_reward_func": 0.0218750003259629,
+      "rewards/thinkcount_reward_func": 0.03906250046566129,
+      "step": 257
+    },
+    {
+      "completion_length": 8117.5,
+      "epoch": 0.01000872854233343,
+      "grad_norm": 5.563852310180664,
+      "kl": 0.01354102953337133,
+      "learning_rate": 4.984484096198604e-07,
+      "loss": 0.0005,
+      "reward": 0.7646483965218067,
+      "reward_std": 0.7928422568365932,
+      "rewards/correctness_reward_func": 0.6787109375,
+      "rewards/soft_format_reward_func": 0.02343750058207661,
+      "rewards/strict_format_reward_func": 0.025000000605359674,
+      "rewards/thinkcount_reward_func": 0.03750000009313226,
+      "step": 258
+    },
+    {
+      "completion_length": 8079.0,
+      "epoch": 0.010047522063815343,
+      "grad_norm": 5.521684169769287,
+      "kl": 0.019811484031379223,
+      "learning_rate": 5.00387897595035e-07,
+      "loss": 0.0008,
+      "reward": 1.1666015088558197,
+      "reward_std": 1.0354629158973694,
+      "rewards/correctness_reward_func": 1.0791015625,
+      "rewards/soft_format_reward_func": 0.02656250074505806,
+      "rewards/strict_format_reward_func": 0.021875000558793545,
+      "rewards/thinkcount_reward_func": 0.039062500931322575,
+      "step": 259
+    },
+    {
+      "completion_length": 8081.5,
+      "epoch": 0.010086315585297255,
+      "grad_norm": 12.548905372619629,
+      "kl": 0.027738861506804824,
+      "learning_rate": 5.023273855702095e-07,
+      "loss": 0.0011,
+      "reward": 1.1030764933675528,
+      "reward_std": 0.8562305881641805,
+      "rewards/correctness_reward_func": 1.0126953125,
+      "rewards/soft_format_reward_func": 0.021875000675208867,
+      "rewards/strict_format_reward_func": 0.03437500121071935,
+      "rewards/thinkcount_reward_func": 0.034131250344216824,
+      "step": 260
     }
   ],
   "logging_steps": 1,