Model save

README.md

@@ -1,11 +1,9 @@
 ---
 base_model: Qwen/Qwen2.5-Coder-7B
-datasets: Genies/Reasoning_SQL_BIRD
 library_name: transformers
 model_name: text2sql-reasoning-sql-no-ex
 tags:
 - generated_from_trainer
-- open-r1
 - trl
 - grpo
 licence: license
@@ -13,7 +11,7 @@ licence: license
 
 # Model Card for text2sql-reasoning-sql-no-ex
 
-This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-7B](https://huggingface.co/Qwen/Qwen2.5-Coder-7B) on the [Genies/Reasoning_SQL_BIRD](https://huggingface.co/datasets/Genies/Reasoning_SQL_BIRD) dataset.
+This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-7B](https://huggingface.co/Qwen/Qwen2.5-Coder-7B).
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start
@@ -29,7 +27,7 @@ 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/genies-rnd/text2sql-rl/runs/rv2w00yx) 
+[<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/genies-rnd/text2sql-rl/runs/v2nly9i2) 
 
 
 This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
@@ -39,7 +37,7 @@ This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing
 - TRL: 0.18.0
 - Transformers: 4.52.3
 - Pytorch: 2.6.0
-- Datasets: 3.6.0
+- Datasets: 4.0.0
 - Tokenizers: 0.21.1
 
 ## Citations

all_results.json

@@ -4,5 +4,11 @@
     "eval_runtime": 1327.7604,
     "eval_samples": 1534,
     "eval_samples_per_second": 1.155,
-    "eval_steps_per_second": 0.024
+    "eval_steps_per_second": 0.024,
+    "total_flos": 0.0,
+    "train_loss": -3.7846087062426205e-06,
+    "train_runtime": 110.1977,
+    "train_samples": 9428,
+    "train_samples_per_second": 256.666,
+    "train_steps_per_second": 8.031
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 0.0003594789403539185,
-    "train_runtime": 4902.9642,
+    "train_loss": -3.7846087062426205e-06,
+    "train_runtime": 110.1977,
     "train_samples": 9428,
-    "train_samples_per_second": 5.769,
-    "train_steps_per_second": 0.181
+    "train_samples_per_second": 256.666,
+    "train_steps_per_second": 8.031
 }

trainer_state.json

@@ -2,9 +2,9 @@
   "best_global_step": null,
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 2.9974554707379135,
+  "epoch": 2.003392705682782,
   "eval_steps": 100,
-  "global_step": 884,
+  "global_step": 885,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
@@ -31834,18 +31834,54 @@
       "step": 884
     },
     {
-      "epoch": 2.9974554707379135,
-      "step": 884,
+      "clip_ratio/high_max": 0.0,
+      "clip_ratio/high_mean": 0.0,
+      "clip_ratio/low_mean": 0.0,
+      "clip_ratio/low_min": 0.0,
+      "clip_ratio/region_mean": 0.0,
+      "completions/clipped_ratio": -1.1875,
+      "completions/max_length": 256.0,
+      "completions/max_terminated_length": 254.0,
+      "completions/mean_length": 222.8072967529297,
+      "completions/mean_terminated_length": 164.95713806152344,
+      "completions/min_length": 50.0,
+      "completions/min_terminated_length": 50.0,
+      "epoch": 2.003392705682782,
+      "frac_reward_zero_std": 0.03125,
+      "grad_norm": 0.7709642648696899,
+      "kl": 0.05810546875,
+      "learning_rate": 3.8941533949121694e-12,
+      "loss": -0.0033,
+      "num_tokens": 226847464.0,
+      "reward": 8.306476593017578,
+      "reward_std": 1.4022259712219238,
+      "rewards/accuracy_reward/mean": 1.171875,
+      "rewards/accuracy_reward/std": 1.4674979448318481,
+      "rewards/format_reward/mean": 0.9479166865348816,
+      "rewards/format_reward/std": 0.22277598083019257,
+      "rewards/judge_reward/mean": 1.0458332300186157,
+      "rewards/judge_reward/std": 0.9318019151687622,
+      "rewards/ngrams_iou_reward/mean": 0.1845172643661499,
+      "rewards/ngrams_iou_reward/std": 0.1875908374786377,
+      "rewards/schema_keywords_iou_reward/mean": 0.7230003476142883,
+      "rewards/schema_keywords_iou_reward/std": 0.18600532412528992,
+      "rewards/syntax_reward/mean": 0.84375,
+      "rewards/syntax_reward/std": 0.36404144763946533,
+      "step": 885
+    },
+    {
+      "epoch": 2.003392705682782,
+      "step": 885,
       "total_flos": 0.0,
-      "train_loss": 0.0003594789403539185,
-      "train_runtime": 4902.9642,
-      "train_samples_per_second": 5.769,
-      "train_steps_per_second": 0.181
+      "train_loss": -3.7846087062426205e-06,
+      "train_runtime": 110.1977,
+      "train_samples_per_second": 256.666,
+      "train_steps_per_second": 8.031
     }
   ],
   "logging_steps": 1,
   "max_steps": 885,
-  "num_input_tokens_seen": 226571285,
+  "num_input_tokens_seen": 226847464,
   "num_train_epochs": 3,
   "save_steps": 52,
   "stateful_callbacks": {
@@ -31855,7 +31891,7 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": false
+        "should_training_stop": true
       },
       "attributes": {}
     }