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-territorial_sleek_impala
 tags:
@@ -8,13 +8,14 @@ tags:
 - grpo
 - gensyn
 - I am territorial sleek impala
+- unsloth
 - trl
 licence: license
 ---
 
 # Model Card for gensyn-checkpoints-territorial_sleek_impala
 
-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",
+    "down_proj",
+    "v_proj",
+    "o_proj",
+    "k_proj",
+    "up_proj",
+    "q_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:ee1a543f126f3f5d98275b0c82f58a26b39028e01c69abdf5f364931d12d172f
+size 73911112

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 4.218931023788173,
-    "train_runtime": 150.0721,
-    "train_samples": 28,
-    "train_samples_per_second": 2.132,
-    "train_steps_per_second": 0.133
+    "train_loss": 8.136904966704605e-07,
+    "train_runtime": 1913.4628,
+    "train_samples": 20,
+    "train_samples_per_second": 0.167,
+    "train_steps_per_second": 0.01
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 4.218931023788173,
-    "train_runtime": 150.0721,
-    "train_samples": 28,
-    "train_samples_per_second": 2.132,
-    "train_steps_per_second": 0.133
+    "train_loss": 8.136904966704605e-07,
+    "train_runtime": 1913.4628,
+    "train_samples": 20,
+    "train_samples_per_second": 0.167,
+    "train_steps_per_second": 0.01
 }

trainer_state.json

@@ -2,7 +2,7 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 5.0,
+  "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": 221.84375,
-      "epoch": 0.5714285714285714,
-      "grad_norm": 34.56943893432617,
-      "kl": 0.0,
+      "completion_length": 417.21875,
+      "epoch": 0.4,
+      "grad_norm": 0.5455081462860107,
+      "kl": 0.0008443489932687953,
       "learning_rate": 5e-07,
       "loss": 0.0,
-      "reward": 3.9288951698690653,
-      "reward_std": 1.1124803191050887,
-      "rewards/concensus_correctness_reward_func": 1.294749990105629,
-      "rewards/consensus_reward_func": 1.0,
+      "reward": 1.2102286852896214,
+      "reward_std": 1.0121423974633217,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.4375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.7205826984718442,
+      "rewards/final_correctness_reward_func": 0.1875,
+      "rewards/question_recreation_reward_func": 0.47594743501394987,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.078125,
-      "rewards/xmlcount_reward_func": 0.8354375129565597,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.07803125120699406,
       "step": 2
     },
     {
-      "completion_length": 180.04166666666666,
-      "epoch": 1.0,
-      "grad_norm": 35.82617950439453,
-      "kl": 0.10452407443275054,
+      "completion_length": 453.5,
+      "epoch": 0.8,
+      "grad_norm": 1.0747790336608887,
+      "kl": 0.0007697510372963734,
       "learning_rate": 4.864543104251586e-07,
-      "loss": 0.0001,
-      "reward": 6.0388092796007795,
-      "reward_std": 0.38356546566744026,
-      "rewards/concensus_correctness_reward_func": 1.939833328127861,
-      "rewards/consensus_reward_func": 1.5,
+      "loss": 0.0,
+      "reward": 1.4177504740655422,
+      "reward_std": 1.4026328399777412,
+      "rewards/concensus_correctness_reward_func": 0.001687500043772161,
+      "rewards/consensus_reward_func": 0.5625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.16666666666666666,
-      "rewards/question_recreation_reward_func": 0.9063508858283361,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6949379779398441,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.375,
-      "rewards/xmlcount_reward_func": 1.1509583344062169,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.14299999782815576,
       "step": 4
     },
     {
-      "completion_length": 200.03125,
-      "epoch": 1.5714285714285714,
-      "grad_norm": 26.39034080505371,
-      "kl": 2.5766645671101287,
+      "completion_length": 366.84375,
+      "epoch": 1.2,
+      "grad_norm": 0.5992124080657959,
+      "kl": 0.000809168501291424,
       "learning_rate": 4.472851273490984e-07,
-      "loss": 0.0026,
-      "reward": 5.809181839227676,
-      "reward_std": 0.4371991759398952,
-      "rewards/concensus_correctness_reward_func": 1.7268749997019768,
-      "rewards/consensus_reward_func": 1.4375,
+      "loss": 0.0,
+      "reward": 1.3867879807949066,
+      "reward_std": 1.3449187986552715,
+      "rewards/concensus_correctness_reward_func": 0.05756250023841858,
+      "rewards/consensus_reward_func": 0.5,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.9328693635761738,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5688504725694656,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.359375,
-      "rewards/xmlcount_reward_func": 1.1650625001639128,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.2447500005364418,
       "step": 6
     },
     {
-      "completion_length": 182.83333333333334,
-      "epoch": 2.0,
-      "grad_norm": 25.702749252319336,
-      "kl": 54736.78004365539,
+      "completion_length": 381.875,
+      "epoch": 1.6,
+      "grad_norm": 0.3977244198322296,
+      "kl": 0.0007945733377709985,
       "learning_rate": 3.867370395306068e-07,
-      "loss": 41.0526,
-      "reward": 5.609723289807637,
-      "reward_std": 0.38764867421317223,
-      "rewards/concensus_correctness_reward_func": 1.5794999798138936,
-      "rewards/consensus_reward_func": 1.4166666666666667,
+      "loss": 0.0,
+      "reward": 1.0844756625592709,
+      "reward_std": 0.8270599953830242,
+      "rewards/concensus_correctness_reward_func": 0.02468749973922968,
+      "rewards/consensus_reward_func": 0.1875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.16666666666666666,
-      "rewards/question_recreation_reward_func": 0.9849732766548792,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.6176319215446711,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.2708333333333333,
-      "rewards/xmlcount_reward_func": 1.1910833368698757,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.25465625058859587,
       "step": 8
     },
     {
-      "completion_length": 207.09375,
-      "epoch": 2.571428571428571,
-      "grad_norm": 45.36337661743164,
-      "kl": 2.440667485818267,
+      "completion_length": 422.0,
+      "epoch": 2.0,
+      "grad_norm": 0.4323352873325348,
+      "kl": 0.0006908488758199383,
       "learning_rate": 3.1137137178519977e-07,
-      "loss": 0.0024,
-      "reward": 5.608146898448467,
-      "reward_std": 0.3853573816741118,
-      "rewards/concensus_correctness_reward_func": 1.5514374934136868,
-      "rewards/consensus_reward_func": 1.5625,
+      "loss": 0.0,
+      "reward": 1.3483711397275329,
+      "reward_std": 1.2780026979744434,
+      "rewards/concensus_correctness_reward_func": 0.039062499068677425,
+      "rewards/consensus_reward_func": 0.625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.9941156804561615,
+      "rewards/final_correctness_reward_func": 0.25,
+      "rewards/question_recreation_reward_func": 0.431089892052114,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.328125,
-      "rewards/xmlcount_reward_func": 1.1719687543809414,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": -0.012406245456077158,
       "step": 10
     },
     {
-      "completion_length": 177.625,
-      "epoch": 3.0,
-      "grad_norm": 4181.13427734375,
-      "kl": 214.9502143630137,
+      "completion_length": 362.71875,
+      "epoch": 2.4,
+      "grad_norm": 0.49867168068885803,
+      "kl": 0.0007399736059596762,
       "learning_rate": 2.2935516363191693e-07,
-      "loss": 0.1612,
-      "reward": 5.501051505406697,
-      "reward_std": 0.4152323539601639,
-      "rewards/concensus_correctness_reward_func": 1.3809999922911327,
-      "rewards/consensus_reward_func": 1.4166666666666667,
+      "loss": 0.0,
+      "reward": 1.1108369305729866,
+      "reward_std": 1.005900513380766,
+      "rewards/concensus_correctness_reward_func": 0.016499999910593033,
+      "rewards/consensus_reward_func": 0.375,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3333333333333333,
-      "rewards/question_recreation_reward_func": 0.9169266273578008,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.49521198123693466,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.3125,
-      "rewards/xmlcount_reward_func": 1.140625,
+      "rewards/strict_format_reward_func": 0.03125,
+      "rewards/xmlcount_reward_func": 0.19287499599158764,
       "step": 12
     },
     {
-      "completion_length": 186.71875,
-      "epoch": 3.571428571428571,
-      "grad_norm": 140.88648986816406,
-      "kl": 308.47393083851784,
+      "completion_length": 314.0,
+      "epoch": 2.8,
+      "grad_norm": 1.8715404272079468,
+      "kl": 0.0008920223699533381,
       "learning_rate": 1.4957614383675767e-07,
-      "loss": 0.3085,
-      "reward": 6.149121820926666,
-      "reward_std": 1.6182817905792035,
-      "rewards/concensus_correctness_reward_func": 2.0923124998807907,
-      "rewards/consensus_reward_func": 1.375,
+      "loss": 0.0,
+      "reward": 1.492088682949543,
+      "reward_std": 1.0026938766241074,
+      "rewards/concensus_correctness_reward_func": -0.004687500186264515,
+      "rewards/consensus_reward_func": 0.5625,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.3125,
-      "rewards/question_recreation_reward_func": 0.9713093787431717,
+      "rewards/final_correctness_reward_func": 0.25,
+      "rewards/question_recreation_reward_func": 0.5516511481255293,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.265625,
-      "rewards/xmlcount_reward_func": 1.1323750019073486,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.11699998192489147,
       "step": 14
     },
     {
-      "completion_length": 160.58333333333334,
-      "epoch": 4.0,
-      "grad_norm": 8.718925476074219,
-      "kl": 1.639391768258065,
+      "completion_length": 304.59375,
+      "epoch": 3.2,
+      "grad_norm": 0.5505232810974121,
+      "kl": 0.0009955409914255142,
       "learning_rate": 8.067960709356478e-08,
-      "loss": 0.0012,
-      "reward": 5.377435078223546,
-      "reward_std": 0.8237529636050264,
-      "rewards/concensus_correctness_reward_func": 1.4001666555802028,
-      "rewards/consensus_reward_func": 1.4166666666666667,
+      "loss": 0.0,
+      "reward": 1.6902727112174034,
+      "reward_std": 0.727166909724474,
+      "rewards/concensus_correctness_reward_func": 0.032999999821186066,
+      "rewards/consensus_reward_func": 0.6875,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.08333333333333333,
-      "rewards/question_recreation_reward_func": 0.9555600434541702,
+      "rewards/final_correctness_reward_func": 0.0,
+      "rewards/question_recreation_reward_func": 0.5042102001607418,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.3333333333333333,
-      "rewards/xmlcount_reward_func": 1.1883749961853027,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.46556251123547554,
       "step": 16
     },
     {
-      "completion_length": 189.34375,
-      "epoch": 4.571428571428571,
-      "grad_norm": 155.6063232421875,
-      "kl": 8.625189328100532,
+      "completion_length": 289.59375,
+      "epoch": 3.6,
+      "grad_norm": 0.6342034339904785,
+      "kl": 0.000844866008264944,
       "learning_rate": 3.013156219837776e-08,
-      "loss": 0.0086,
-      "reward": 5.694100014865398,
-      "reward_std": 0.40210114730143687,
-      "rewards/concensus_correctness_reward_func": 1.6176249831914902,
-      "rewards/consensus_reward_func": 1.625,
+      "loss": 0.0,
+      "reward": 1.9497819244861603,
+      "reward_std": 1.0366131216287613,
+      "rewards/concensus_correctness_reward_func": 0.0,
+      "rewards/consensus_reward_func": 0.75,
       "rewards/cumulative_reward_2": 0.0,
-      "rewards/final_correctness_reward_func": 0.1875,
-      "rewards/question_recreation_reward_func": 0.9089125599712133,
+      "rewards/final_correctness_reward_func": 0.125,
+      "rewards/question_recreation_reward_func": 0.6218756604939699,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.265625,
-      "rewards/xmlcount_reward_func": 1.0894374996423721,
+      "rewards/strict_format_reward_func": 0.0,
+      "rewards/xmlcount_reward_func": 0.4529062481597066,
       "step": 18
     },
     {
-      "completion_length": 172.16666666666666,
-      "epoch": 5.0,
-      "grad_norm": 57570.94140625,
-      "kl": 869.4708892249813,
+      "completion_length": 404.3125,
+      "epoch": 4.0,
+      "grad_norm": 0.7241014838218689,
+      "kl": 0.0007424046125379391,
       "learning_rate": 3.4096741493194193e-09,
-      "loss": 0.6521,
-      "reward": 4.955635358889897,
-      "reward_std": 0.4001488390301044,
-      "rewards/concensus_correctness_reward_func": 1.3370000024636586,
-      "rewards/consensus_reward_func": 1.0833333333333333,
+      "loss": 0.0,
+      "reward": 1.7487756162881851,
+      "reward_std": 1.3057930991053581,
+      "rewards/concensus_correctness_reward_func": 0.04499999922700226,
+      "rewards/consensus_reward_func": 0.875,
       "rewards/cumulative_reward_2": 0.0,
       "rewards/final_correctness_reward_func": 0.0,
-      "rewards/question_recreation_reward_func": 0.9936353713274002,
+      "rewards/question_recreation_reward_func": 0.5349631533026695,
       "rewards/soft_format_reward_func": 0.0,
-      "rewards/strict_format_reward_func": 0.3541666666666667,
-      "rewards/xmlcount_reward_func": 1.1875,
+      "rewards/strict_format_reward_func": 0.015625,
+      "rewards/xmlcount_reward_func": 0.27818750124424696,
       "step": 20
     },
     {
-      "epoch": 5.0,
+      "epoch": 4.0,
       "step": 20,
       "total_flos": 0.0,
-      "train_loss": 4.218931023788173,
-      "train_runtime": 150.0721,
-      "train_samples_per_second": 2.132,
-      "train_steps_per_second": 0.133
+      "train_loss": 8.136904966704605e-07,
+      "train_runtime": 1913.4628,
+      "train_samples_per_second": 0.167,
+      "train_steps_per_second": 0.01
     }
   ],
   "logging_steps": 2,
   "max_steps": 20,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 7,
+  "num_train_epochs": 4,
   "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:f3ab5140359ad9477026a3086757538e96dc8b269a9f61270fc58f7b93643ea8
-size 5880
+oid sha256:ed90e4452f86665212b8c15189b1ac85f7ce1e368cd29efcdac71b9835ad1e7f
+size 5944