Upload README.md with huggingface_hub

README.md

@@ -1,6 +1,20 @@
+---
+pipeline_tag: text-generation
+library_name: transformers
+tags:
+- mining
+- russian
+license: apache-2.0
+datasets:
+- nn-tech/Alloy-Bench
+language:
+- ru
+base_model: Qwen/Qwen3-32B
+---
+
 ## Description
 
-**MetalGPT-1** is a model built upon the Qwen3-32b and incorporates both continual pre-training and supervised fine-tuning.
+**MetalGPT-1** is a model built upon the Qwen/Qwen3-32B and incorporates both continual pre-training and supervised fine-tuning.
 
 ---
 
@@ -43,7 +57,7 @@ model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
 
 generated_ids = model.generate(
     **model_inputs,
-    max_new_tokens=1024,
+    max_new_tokens=2048,
 )
 
 # Обрезаем префикс промпта
@@ -132,7 +146,7 @@ sampling_params = SamplingParams(
     repetition_penalty=1.05,
     top_p=0.8,
     top_k=70,
-    max_tokens=512,
+    max_tokens=2048,
 )
 
 prompt = (