End of training

README.md

@@ -1,5 +1,5 @@
 ---
-base_model: Gensyn/Qwen2.5-0.5B-Instruct
+base_model: Gensyn/Qwen2.5-1.5B-Instruct
 library_name: transformers
 model_name: gensyn-checkpoints-soaring_polished_cobra
 tags:
@@ -8,13 +8,14 @@ tags:
 - grpo
 - gensyn
 - I am soaring polished cobra
+- unsloth
 - trl
 licence: license
 ---
 
 # Model Card for gensyn-checkpoints-soaring_polished_cobra
 
-This model is a fine-tuned version of [Gensyn/Qwen2.5-0.5B-Instruct](https://huggingface.co/Gensyn/Qwen2.5-0.5B-Instruct).
+This model is a fine-tuned version of [Gensyn/Qwen2.5-1.5B-Instruct](https://huggingface.co/Gensyn/Qwen2.5-1.5B-Instruct).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
@@ -39,7 +40,7 @@ This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing
 
 - TRL: 0.15.2
 - Transformers: 4.51.3
-- Pytorch: 2.5.1
+- Pytorch: 2.6.0
 - Datasets: 3.5.0
 - Tokenizers: 0.21.1
 

adapter_config.json

@@ -0,0 +1,39 @@
+{
+  "alpha_pattern": {},
+  "auto_mapping": null,
+  "base_model_name_or_path": "Gensyn/Qwen2.5-1.5B-Instruct",
+  "bias": "none",
+  "corda_config": null,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 16,
+  "lora_bias": false,
+  "lora_dropout": 0,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "gate_proj",
+    "q_proj",
+    "o_proj",
+    "k_proj",
+    "down_proj",
+    "up_proj",
+    "v_proj"
+  ],
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ef0c9fc2a22680be6b9b8fa9540f1875dc6c049693303b0d3390ad2c07c40f9d
+size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.716044910921482e-06,
-    "train_runtime": 943.9196,
-    "train_samples": 51,
-    "train_samples_per_second": 0.339,
-    "train_steps_per_second": 0.021
+    "train_loss": 6.641451051336844e-07,
+    "train_runtime": 718.6924,
+    "train_samples": 17,
+    "train_samples_per_second": 0.445,
+    "train_steps_per_second": 0.028
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 7.716044910921482e-06,
-    "train_runtime": 943.9196,
-    "train_samples": 51,
-    "train_samples_per_second": 0.339,
-    "train_steps_per_second": 0.021
+    "train_loss": 6.641451051336844e-07,
+    "train_runtime": 718.6924,
+    "train_samples": 17,
+    "train_samples_per_second": 0.445,
+    "train_steps_per_second": 0.028
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.9411764705882355,
+  "epoch": 4.0,
   "eval_steps": 500,
   "global_step": 20,
   "is_hyper_param_search": false,
@@ -10,209 +10,209 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "completion_length": 320.21875,
-      "epoch": 0.3137254901960784,
-      "grad_norm": 18.987831115722656,
-      "kl": 0.0,
+      "completion_length": 427.875,
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.5876138806343079,
+      "kl": 0.0007815321441739798,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 0.9052386409603059,
-      "reward_std": 1.0872065802104771,
-      "rewards/concensus_correctness_reward_func": 0.0078125,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.6895362362265587,
+      "reward_std": 1.607804922387004,
+      "rewards/concensus_correctness_reward_func": 0.30674999952316284,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0625,
-      "rewards/question_recreation_reward_func": 0.32886362238787115,
-      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/final_correctness_reward_func": 0.3125,
+      "rewards/question_recreation_reward_func": 0.5297549953684211,
+      "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.22481249924749136,
+      "rewards/xmlcount_reward_func": 0.14990626135841012,
       "step": 2
     },
     {
-      "completion_length": 255.65625,
-      "epoch": 0.6274509803921569,
-      "grad_norm": 7.594056606292725,
-      "kl": 0.0013956625989521854,
+      "completion_length": 340.0,
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.98947674036026,
+      "kl": 0.0008097236932371743,
       "learning_rate": 4.864543104251586e-07,
       "loss": 0.0,
-      "reward": 1.7687878240831196,
-      "reward_std": 0.8872564652119763,
-      "rewards/concensus_correctness_reward_func": 0.19506250228732824,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 1.1686192080378532,
+      "reward_std": 0.8920873831957579,
+      "rewards/concensus_correctness_reward_func": 0.0078125,
+      "rewards/consensus_reward_func": 0.25,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.5405378290452063,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.6581875095143914,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.562056683935225,
+      "rewards/soft_format_reward_func": 0.03125,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.301875002682209,
       "step": 4
     },
     {
-      "completion_length": 241.71875,
-      "epoch": 0.9411764705882353,
-      "grad_norm": 5.940536022186279,
-      "kl": 0.0032457121487823315,
+      "completion_length": 538.9,
+      "epoch": 1.2352941176470589,
+      "grad_norm": 0.518266499042511,
+      "kl": 0.0006855448475107551,
       "learning_rate": 4.472851273490984e-07,
       "loss": 0.0,
-      "reward": 3.927860476076603,
-      "reward_std": 1.7695390642620623,
-      "rewards/concensus_correctness_reward_func": 2.1019999980926514,
-      "rewards/consensus_reward_func": 0.375,
+      "reward": 2.2191823825240133,
+      "reward_std": 1.331932282447815,
+      "rewards/concensus_correctness_reward_func": 0.525,
+      "rewards/consensus_reward_func": 0.9,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.43264187331078574,
+      "rewards/final_correctness_reward_func": 0.3,
+      "rewards/question_recreation_reward_func": 0.5259323239326477,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.7682187557220459,
+      "rewards/xmlcount_reward_func": -0.031749993562698364,
       "step": 6
     },
     {
-      "completion_length": 256.0,
-      "epoch": 1.156862745098039,
-      "grad_norm": 9.757232666015625,
-      "kl": 0.00435605255717581,
+      "completion_length": 329.9375,
+      "epoch": 1.7058823529411766,
+      "grad_norm": 0.4131653308868408,
+      "kl": 0.0007813851116225123,
       "learning_rate": 3.867370395306068e-07,
       "loss": 0.0,
-      "reward": 1.873472045768391,
-      "reward_std": 0.48061720840632915,
-      "rewards/concensus_correctness_reward_func": 0.15699999982660467,
-      "rewards/consensus_reward_func": 0.09090909090909091,
+      "reward": 1.6284910812973976,
+      "reward_std": 0.8130084574222565,
+      "rewards/concensus_correctness_reward_func": 0.11649999767541885,
+      "rewards/consensus_reward_func": 0.3125,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.2727272727272727,
-      "rewards/question_recreation_reward_func": 0.7095629688013684,
+      "rewards/final_correctness_reward_func": 0.4375,
+      "rewards/question_recreation_reward_func": 0.4998348420485854,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.643272725018588,
+      "rewards/xmlcount_reward_func": 0.26215624529868364,
       "step": 8
     },
     {
-      "completion_length": 158.59375,
-      "epoch": 1.4705882352941178,
-      "grad_norm": 6.622710227966309,
-      "kl": 0.010859893664019182,
+      "completion_length": 403.2,
+      "epoch": 2.0,
+      "grad_norm": 0.13836334645748138,
+      "kl": 0.0006139612174592912,
       "learning_rate": 3.1137137178519977e-07,
       "loss": 0.0,
-      "reward": 3.0381277231499553,
-      "reward_std": 0.5335058589698747,
-      "rewards/concensus_correctness_reward_func": 1.4662500023841858,
-      "rewards/consensus_reward_func": 0.4375,
+      "reward": 2.0887853801250458,
+      "reward_std": 0.6115315139293671,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 1.1,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.36175276967696846,
+      "rewards/question_recreation_reward_func": 0.5645354151725769,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.77262500859797,
+      "rewards/xmlcount_reward_func": 0.4242500215768814,
       "step": 10
     },
     {
-      "completion_length": 283.375,
-      "epoch": 1.784313725490196,
-      "grad_norm": 5.638906002044678,
-      "kl": 0.007406034710584208,
+      "completion_length": 444.5625,
+      "epoch": 2.4705882352941178,
+      "grad_norm": 0.40184539556503296,
+      "kl": 0.0006372530478984118,
       "learning_rate": 2.2935516363191693e-07,
       "loss": 0.0,
-      "reward": 1.6740143587812781,
-      "reward_std": 1.191217671148479,
-      "rewards/concensus_correctness_reward_func": 0.1471250019967556,
-      "rewards/consensus_reward_func": 0.1875,
+      "reward": 1.8296705558896065,
+      "reward_std": 1.1184727065265179,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.25,
-      "rewards/question_recreation_reward_func": 0.48260811483487487,
+      "rewards/final_correctness_reward_func": 0.0625,
+      "rewards/question_recreation_reward_func": 0.5306080486625433,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.6067812591791153,
+      "rewards/xmlcount_reward_func": 0.36156250536441803,
       "step": 12
     },
     {
-      "completion_length": 249.0,
-      "epoch": 2.0,
-      "grad_norm": 3.3914954662323,
-      "kl": 0.011619402540170333,
+      "completion_length": 349.3125,
+      "epoch": 2.9411764705882355,
+      "grad_norm": 0.648057758808136,
+      "kl": 0.0008520848932676017,
       "learning_rate": 1.4957614383675767e-07,
       "loss": 0.0,
-      "reward": 4.049619214101271,
-      "reward_std": 2.009789251807061,
-      "rewards/concensus_correctness_reward_func": 2.756545454263687,
-      "rewards/consensus_reward_func": 0.09090909090909091,
+      "reward": 1.988405354321003,
+      "reward_std": 1.3347271010279655,
+      "rewards/concensus_correctness_reward_func": 0.2265625,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.09090909090909091,
-      "rewards/question_recreation_reward_func": 0.4135735107755119,
+      "rewards/final_correctness_reward_func": 0.3125,
+      "rewards/question_recreation_reward_func": 0.5402803681790829,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.022727272727272728,
-      "rewards/xmlcount_reward_func": 0.6749545498327776,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.37781250290572643,
       "step": 14
     },
     {
-      "completion_length": 223.875,
-      "epoch": 2.313725490196078,
-      "grad_norm": 8.215188026428223,
-      "kl": 0.014186236599925905,
+      "completion_length": 408.5,
+      "epoch": 3.235294117647059,
+      "grad_norm": 0.6705062389373779,
+      "kl": 0.0009508899878710508,
       "learning_rate": 8.067960709356478e-08,
       "loss": 0.0,
-      "reward": 2.642803938128054,
-      "reward_std": 1.4382841207552701,
-      "rewards/concensus_correctness_reward_func": 0.8331250017508864,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.2143768668174744,
+      "reward_std": 0.8865768074989319,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.4,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3125,
-      "rewards/question_recreation_reward_func": 0.4023352946387604,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6206269025802612,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.015625,
-      "rewards/xmlcount_reward_func": 0.829218752682209,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.19375,
       "step": 16
     },
     {
-      "completion_length": 247.875,
-      "epoch": 2.627450980392157,
-      "grad_norm": 7.052781105041504,
-      "kl": 0.015200054476736113,
+      "completion_length": 320.40625,
+      "epoch": 3.7058823529411766,
+      "grad_norm": 1.4544986486434937,
+      "kl": 0.0009329580789199099,
       "learning_rate": 3.013156219837776e-08,
       "loss": 0.0,
-      "reward": 1.8882133960723877,
-      "reward_std": 1.274847088381648,
-      "rewards/concensus_correctness_reward_func": 0.06793750077486038,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 2.1640911400318146,
+      "reward_std": 1.306618556380272,
+      "rewards/concensus_correctness_reward_func": 0.27806249633431435,
+      "rewards/consensus_reward_func": 0.625,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.3125,
-      "rewards/question_recreation_reward_func": 0.5125571433454752,
-      "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.046875,
-      "rewards/xmlcount_reward_func": 0.698343763127923,
+      "rewards/question_recreation_reward_func": 0.5214036731049418,
+      "rewards/soft_format_reward_func": 0.015625,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.3802500069141388,
       "step": 18
     },
     {
-      "completion_length": 253.28125,
-      "epoch": 2.9411764705882355,
-      "grad_norm": 7.9368438720703125,
-      "kl": 0.013815053927828558,
+      "completion_length": 395.4,
+      "epoch": 4.0,
+      "grad_norm": 0.2986851930618286,
+      "kl": 0.0007005963008850813,
       "learning_rate": 3.4096741493194193e-09,
       "loss": 0.0,
-      "reward": 3.0817041154950857,
-      "reward_std": 2.74389247270301,
-      "rewards/concensus_correctness_reward_func": 1.3802500003948808,
-      "rewards/consensus_reward_func": 0.25,
+      "reward": 1.7007000684738158,
+      "reward_std": 0.8157501250505448,
+      "rewards/concensus_correctness_reward_func": 0.028499999642372133,
+      "rewards/consensus_reward_func": 0.9,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.48770411510486156,
+      "rewards/final_correctness_reward_func": 0.2,
+      "rewards/question_recreation_reward_func": 0.4101000681519508,
       "rewards/soft_format_reward_func": 0.0,
       "rewards/strict_format_reward_func": 0.0,
-      "rewards/xmlcount_reward_func": 0.7762499954551458,
+      "rewards/xmlcount_reward_func": 0.16210000067949296,
       "step": 20
     },
     {
-      "epoch": 2.9411764705882355,
+      "epoch": 4.0,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 7.716044910921482e-06,
-      "train_runtime": 943.9196,
-      "train_samples_per_second": 0.339,
-      "train_steps_per_second": 0.021
+      "train_loss": 6.641451051336844e-07,
+      "train_runtime": 718.6924,
+      "train_samples_per_second": 0.445,
+      "train_steps_per_second": 0.028
     }
   ],
   "logging_steps": 2,
   "max_steps": 20,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 5,
   "save_steps": 25,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -227,7 +227,7 @@
     }
   },
   "total_flos": 0.0,
-  "train_batch_size": 2,
+  "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null
 }

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:44ab3cb72ef003efdc56d9938ef1b19649d4088db0b36611fd01fbeb752fdfc4
-size 5880
+oid sha256:60d32258253d459158070110279c739eea3792bcfbf0af7487e04a785c458a2f
+size 5944