End of training

README.md

@@ -1,9 +1,11 @@
 ---
 base_model: sai-lakkshmii/Finetuned-llama-8b_SFT_step5000
+datasets: DigitalLearningGmbH/MATH-lighteval
 library_name: transformers
 model_name: Final-try-Llama3.1-8b-instruct-RL
 tags:
 - generated_from_trainer
+- open-r1
 - trl
 - grpo
 licence: license
@@ -11,7 +13,7 @@ licence: license
 
 # Model Card for Final-try-Llama3.1-8b-instruct-RL
 
-This model is a fine-tuned version of [sai-lakkshmii/Finetuned-llama-8b_SFT_step5000](https://huggingface.co/sai-lakkshmii/Finetuned-llama-8b_SFT_step5000).
+This model is a fine-tuned version of [sai-lakkshmii/Finetuned-llama-8b_SFT_step5000](https://huggingface.co/sai-lakkshmii/Finetuned-llama-8b_SFT_step5000) on the [DigitalLearningGmbH/MATH-lighteval](https://huggingface.co/datasets/DigitalLearningGmbH/MATH-lighteval) dataset.
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start

all_results.json

@@ -1,4 +1,9 @@
 {
+    "eval_loss": -0.16441404819488525,
+    "eval_runtime": 5255.7962,
+    "eval_samples": 5000,
+    "eval_samples_per_second": 0.951,
+    "eval_steps_per_second": 0.009,
     "total_flos": 0.0,
     "train_loss": -0.0886706564703892,
     "train_runtime": 9867.8585,

config.json

@@ -35,6 +35,6 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.49.0",
-  "use_cache": false,
+  "use_cache": true,
   "vocab_size": 128256
 }

eval_results.json

@@ -0,0 +1,7 @@
+{
+    "eval_loss": -0.16441404819488525,
+    "eval_runtime": 5255.7962,
+    "eval_samples": 5000,
+    "eval_samples_per_second": 0.951,
+    "eval_steps_per_second": 0.009
+}