This model is a fine-tuned version of deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B, enhanced with chain-of-thought (CoT) reasoning on a hybrid dataset combining GSM8K-style reasoning and structured text-to-SQL generation.

Fine-tuning was conducted using DeepSpeed on a multi-A100 GPU setup via RunPod for efficient training in memory-constrained environments. The training dataset includes complex logical SQL queries generated synthetically with corresponding natural language prompts and CoT explanations.

For inference, please see this publicly available notebook.

🧾 Model Details

Base Model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
Language: English
Architecture: Causal Language Model (Decoder-only)
Tokenizer: AutoTokenizer from base model
Parameter Count: 1.5 Billion
Training Framework: 🤗 Transformers + DeepSpeed
Compute Environment: RunPod (6x A100 SXM, 192 vCPU, 1.5TB RAM)

🧪 Training Dataset

Dataset Used:

The dataset contains structured training examples of the form:

xml

<question>...</question>
<think>...</think>
<response>...</response>

Each example is constructed to include:

A natural language question about tabular data (sql_prompt)
An intermediate reasoning step (sql_explanation)
The final SQL output (sql)

This format allows the model to internalize step-by-step logical reasoning for SQL generation.

📊 Fine-Tuning Summary

The base model deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B was fine-tuned on three different datasets using DeepSpeed across various RunPod infrastructure setups. Below is a consolidated summary of the training configurations and results:

Model ID	Dataset Description	GPUs	vCPUs	RAM (GB)	Disk per GPU	Container Image	Duration	Cost	DeepSpeed Stage	Precision	Mean Token Accuracy
`eagle0504/finetuned-deepseek-r1-distill-qwen-1.5b-by-openai-gsm8k-enhanced-v2`	OpenAI GSM8K Enhanced v2	6 × H100 PCIe	144	1132	20 GB	`runpod/pytorch:2.1.0-py3.10-cuda11.8.0-devel-ubuntu22.04`	2 hrs	~$28	Stage 1	FP16	98%
`eagle0504/openai-gsm8k-codealpaca-20k-enhanced-deepseek-r1-distill-qwen-1.5b`	GSM8K + CodeAlpaca-20K Enhanced	4 × A100 SXM	146	1144	20 GB	`runpod/pytorch:2.1.0-py3.10-cuda11.8.0-devel-ubuntu22.04`	2 hrs	~$14+	Stage 1	FP16	97%
`eagle0504/qwen-distilled-scout-1.5b`	Custom CoT + SQL-Reasoning	6 × A100 SXM	192	1536	20 GB	`runpod/pytorch:2.1.0-py3.10-cuda11.8.0-devel-ubuntu22.04`	1.5 hrs	~$21	Stage 2	FP16	97%

🧮 Evaluation Metric

The model is evaluated with a custom token-level accuracy metric:

Metric: Mean token-level accuracy
Definition: Accuracy over all non-masked tokens (labels != -100)
Implementation: NumPy-based vectorized comparison between predicted tokens and ground truth

🚀 Use Case

The model is designed for chain-of-thought text-to-SQL generation, useful in:

AI teaching agents
Conversational agents with data query capabilities
Auto SQL generation tools for tabular backends
Educational applications in logical reasoning

📦 How to Use

python

from transformers import StoppingCriteria, StoppingCriteriaList
import torch

class StopOnTokens(StoppingCriteria):
    def __init__(self, stop_token_ids: list):
        super().__init__()
        self.stop_token_ids = stop_token_ids

    def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs) -> bool:
        # Check if the last token matches any of the stop tokens
        return any(input_ids[0, -len(token):].tolist() == token for token in self.stop_token_ids)

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained("eagle0504/qwen-distilled-scout-1.5b-gen2")
tokenizer = AutoTokenizer.from_pretrained("eagle0504/qwen-distilled-scout-1.5b-gen2")

# Example stop sequence
stop_sequence = "</response>"
stop_ids = tokenizer.encode(stop_sequence, add_special_tokens=False)
stopping_criteria = StoppingCriteriaList([StopOnTokens([stop_ids])])

# Run generation with stop sequence
inputs = tokenizer(
    "<question>Natalia sold clips to 48 of her friends in April, and then she sold half as many clips in May. How many clips did Natalia sell altogether in April and May?</question>",
    return_tensors="pt"
)

outputs = model.generate(
    **inputs,
    max_new_tokens=230,
    stopping_criteria=stopping_criteria
)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))

📊 Limitations

The model is tuned for text-to-SQL tasks with CoT supervision and may not generalize well to free-form text generation or other domains without additional fine-tuning.
Maximum input length is 1024 tokens — longer contexts will be truncated.

🧑‍💻 Author

Name: Yiqiao Yin
Hugging Face: eagle0504
Organization: [WYN AI / Independent AI Researcher]

📝 Citation

If you use this model in your work, please cite:

bibtex

@misc{yin2025enhanceddeepseek,
  title={Enhanced DeepSeek-R1-Distill-Qwen-1.5B Fine-tuned on GSM8K + CoT SQL},
  author={Yiqiao Yin},
  year={2025},
  howpublished={\url{https://huggingface.co/eagle0504/enhanced-deepseek-r1-distill-qwen-1.5b-finetuned-on-gsm8k-codealpaca20k-text2sql}},
}

📬 Contact

For questions or collaborations, reach out via LinkedIn or email: eagle0504@gmail.com