iamgiven/Qwen2.5-Coder-1.5B-Instruct-cpp

1 2---3language:4- en5- id6tags:7- qwen8- code9- merged10- optimized11pipeline_tag: text-generation12license: apache-2.013---14 15# Qwen2.5 Coder 1.5B Instruct Merged (Optimized)16 17This is an optimized merged version of the fine-tuned Qwen2.5 Coder model. It combines:18- Base model: Qwen/Qwen2.5-Coder-1.5B-Instruct19- Fine-tuned adapter: iamgiven/Qwen2.5-Coder-1.5B-Instruct-cpp-lora20 21The model has been optimized using float16 precision and efficient serialization.22 23## Usage24 25```python26from transformers import AutoModelForCausalLM, AutoTokenizer27import torch28 29model = AutoModelForCausalLM.from_pretrained(30    "{full_repo_name}",31    trust_remote_code=True,32    torch_dtype=torch.float16  # Use float16 for efficiency33)34tokenizer = AutoTokenizer.from_pretrained("{full_repo_name}", trust_remote_code=True)35```36