End of training

README.md

@@ -1,9 +1,21 @@
 ---
 base_model: Qwen/Qwen3-32B
+datasets:
+- neko-llm/HLE_SFT_general
+- neko-llm/HLE_SFT_OpenThoughts-114k
+- neko-llm/HLE_SFT_OpenMathReasoning
+- neko-llm/HLE_SFT_MixtureOfThoughts
+- neko-llm/HLE_SFT_MixtureOfThoughts
+- neko-llm/HLE_SFT_Chemistry
+- neko-llm/HLE_SFT_biology
+- neko-llm/HLE_SFT_medical
+- neko-llm/HLE_SFT_general
+- neko-llm/HLE_SFT_humanity
 library_name: transformers
 model_name: Qwen3-32B-HLE
 tags:
 - generated_from_trainer
+- open-r1
 - sft
 - trl
 licence: license
@@ -11,7 +23,7 @@ licence: license
 
 # Model Card for Qwen3-32B-HLE
 
-This model is a fine-tuned version of [Qwen/Qwen3-32B](https://huggingface.co/Qwen/Qwen3-32B).
+This model is a fine-tuned version of [Qwen/Qwen3-32B](https://huggingface.co/Qwen/Qwen3-32B) on the [['neko-llm/HLE_SFT_general', 'neko-llm/HLE_SFT_OpenThoughts-114k', 'neko-llm/HLE_SFT_OpenMathReasoning', 'neko-llm/HLE_SFT_MixtureOfThoughts', 'neko-llm/HLE_SFT_MixtureOfThoughts', 'neko-llm/HLE_SFT_Chemistry', 'neko-llm/HLE_SFT_biology', 'neko-llm/HLE_SFT_medical', 'neko-llm/HLE_SFT_general', 'neko-llm/HLE_SFT_humanity']](https://huggingface.co/datasets/['neko-llm/HLE_SFT_general', 'neko-llm/HLE_SFT_OpenThoughts-114k', 'neko-llm/HLE_SFT_OpenMathReasoning', 'neko-llm/HLE_SFT_MixtureOfThoughts', 'neko-llm/HLE_SFT_MixtureOfThoughts', 'neko-llm/HLE_SFT_Chemistry', 'neko-llm/HLE_SFT_biology', 'neko-llm/HLE_SFT_medical', 'neko-llm/HLE_SFT_general', 'neko-llm/HLE_SFT_humanity']) dataset.
 It has been trained using [TRL](https://github.com/huggingface/trl).
 
 ## Quick start

config.json

@@ -90,7 +90,7 @@
   "tie_word_embeddings": false,
   "torch_dtype": "bfloat16",
   "transformers_version": "4.54.1",
-  "use_cache": false,
+  "use_cache": true,
   "use_sliding_window": false,
   "vocab_size": 151936
 }