Model save

README.md

@@ -1,11 +1,9 @@
 ---
-base_model: Genies/text2sql-sft-kumar-v4
-datasets: Genies/text2sql-grpo-d6
+base_model: Qwen/Qwen2.5-Coder-14B-Instruct
 library_name: transformers
 model_name: text2sql-grpo-intermediate-reward
 tags:
 - generated_from_trainer
-- open-r1
 - trl
 - grpo
 licence: license
@@ -13,7 +11,7 @@ licence: license
 
 # Model Card for text2sql-grpo-intermediate-reward
 
-This model is a fine-tuned version of [Genies/text2sql-sft-kumar-v4](https://huggingface.co/Genies/text2sql-sft-kumar-v4) on the [Genies/text2sql-grpo-d6](https://huggingface.co/datasets/Genies/text2sql-grpo-d6) dataset.
+This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-14B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-14B-Instruct).
 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/9f5vruq2) 
+[<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/ii9plj8l) 
 
 
 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).

all_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 0.025686448466030635,
-    "train_runtime": 272293.2478,
+    "train_loss": 0.002633511937032877,
+    "train_runtime": 233060.1474,
     "train_samples": 14159,
-    "train_samples_per_second": 0.104,
-    "train_steps_per_second": 0.002
+    "train_samples_per_second": 0.122,
+    "train_steps_per_second": 0.001
 }

generation_config.json

@@ -6,7 +6,7 @@
     151643
   ],
   "pad_token_id": 151643,
-  "repetition_penalty": 1.1,
+  "repetition_penalty": 1.05,
   "temperature": 0.7,
   "top_k": 20,
   "top_p": 0.8,

train_results.json

@@ -1,8 +1,8 @@
 {
     "total_flos": 0.0,
-    "train_loss": 0.025686448466030635,
-    "train_runtime": 272293.2478,
+    "train_loss": 0.002633511937032877,
+    "train_runtime": 233060.1474,
     "train_samples": 14159,
-    "train_samples_per_second": 0.104,
-    "train_steps_per_second": 0.002
+    "train_samples_per_second": 0.122,
+    "train_steps_per_second": 0.001
 }