- **Finetuned from model [optional]: Qwen/Qwen2.5-Coder-7B-Instruct ** [More Information Needed]
How to Get Started with the Model
Use the code below to get started with the model.
from transformers import AutoTokenizer, AutoModelForCausalLM
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
model_name = "huge-michael/sylvan-model"
This automatically downloads the model and tokenizer files
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True, device_map="auto")
model.eval()
Define your test query
prompt = """
write a function def process_and_analyze_csv_data(df) to:
Processes and analyzes a CSV dataset for train-test split, feature scaling, and data statistics.
The function should output with:
dict: A dictionary containing train/test split data, scaled features, and statistical data.
You should start with:
['pandas as pd', 'numpy as np', 'train_test_split', 'StandardScaler']
def process_and_analyze_csv_data(df)"""
Construct Qwen-style messages
messages = [
{"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."},
{"role": "user", "content": prompt}
]
Convert the messages into the Qwen prompt format
text = tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True # adds the Assistant role placeholder for generation
)
Tokenize the prompt
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
Generate the response
with torch.no_grad():
outputs = model.generate(
**model_inputs,
max_new_tokens=800,
temperature=0.7, # adjust for randomness
top_p=0.9 # adjust for response diversity
)
Remove the prompt tokens from the generation to isolate the assistant’s response
generated_ids = [
output_ids[len(input_ids):]
for input_ids, output_ids in zip(model_inputs.input_ids, outputs)
]
Decode the generated tokens
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
print("Assistant response:\n", response)
[More Information Needed]
