TabRankStandardSFT

TabRankStandardSFT

Qwen3-8B, fine-tuned for single-call generative listwise table reranking. Given a question and a list of candidate tables, it reads them all in one prompt and returns the full ranking in a single generation — no pairwise scoring, no cross-encoder passes.

This checkpoint is trained with standard chain-of-thought distillation: loss over the full reasoning trace plus the final ranking, distilled from 6,728 teacher reasoning traces. It writes its own <think>...</think> reasoning at inference. It generalizes less reliably out-of-distribution than TabRank, our reasoning-conditioned method — see the paper for the comparison and ablations.

Related: TabRank (same data, reasoning-conditioned, our best method) · TabRankNaive (same data, no reasoning trace, fastest).

Input / output format

Input — a chat message with the question followed by each candidate table, labeled ### Table 1, ### Table 2, ...:

Question: Which table shows 2022 quarterly revenue by region?

### Table 1
| Region | Q1 2022 | Q2 2022 | Q3 2022 | Q4 2022 |
|---|---|---|---|---|
| North America | 120 | 134 | 128 | 145 |
| Europe | 88 | 91 | 95 | 102 |

### Table 2
| Product | Units Sold | Year |
|---|---|---|
| Widget A | 4200 | 2021 |

### Table 3
| Region | Headcount |
|---|---|
| North America | 340 |

Output — a <think> block with the model's reasoning, followed by a single JSON object with the ranked, one-indexed candidate positions, best first:

<think>
Table 1 has quarterly revenue by region for 2022, which is exactly what the question asks
for. Table 3 has region data but no revenue. Table 2 has neither region nor 2022 data.
</think>
{"ranked_tables": [1, 3, 2]}

Map the numbers back to your own table ids to get the reranked list — position 1 in the output is ### Table 1 from the input, etc.

Evaluation

Scored as a listwise reranker reordering a first-stage top-25 candidate list on 5 in-distribution benchmarks (SQA, TAT-QA, HybridQA, TabFact, and NQ-Tables — the actual training split) and 7 out-of-distribution benchmarks from the IBM table-text-ir-evaluation suite that this model never saw during training.

nDCG@10. The first 5 columns are in-distribution; the remaining 7 are out-of-distribution. This checkpoint sits between Base and TabRank on mean nDCG@10 — modest gains from CoT distillation, but weaker generalization and ~4.3x slower inference than the reasoning-conditioned TabRank method.

Full eval code and logs: GitHub.

Usage with vLLM

from vllm import LLM, SamplingParams
from transformers import AutoTokenizer

repo = "AdarshSingh7647/TabRankStandardSFT"
tok = AutoTokenizer.from_pretrained(repo)
llm = LLM(model=repo, dtype="bfloat16", max_model_len=32768)

system = ("You are a table relevance expert. Given a question and a set of candidate tables "
          "rank them from most to least useful for answering the question. Reason in a "
          "<think>...</think> block then output exactly JSON with key ranked_tables.")
user = "Question: ...\n\n### Table 1\n...\n\n### Table 2\n...\n"

msgs = [{"role": "system", "content": system}, {"role": "user", "content": user}]
text = tok.apply_chat_template(msgs, tokenize=False, add_generation_prompt=True)
out = llm.generate([text], SamplingParams(temperature=0.6, top_p=0.95, max_tokens=8192))
print(out[0].outputs[0].text)

Usage with Transformers

import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

repo = "AdarshSingh7647/TabRankStandardSFT"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(repo, torch_dtype=torch.bfloat16, device_map="auto")

system = ("You are a table relevance expert. Given a question and a set of candidate tables "
          "rank them from most to least useful for answering the question. Reason in a "
          "<think>...</think> block then output exactly JSON with key ranked_tables.")
user = "Question: ...\n\n### Table 1\n...\n\n### Table 2\n...\n"

msgs = [{"role": "system", "content": system}, {"role": "user", "content": user}]
text = tok.apply_chat_template(msgs, tokenize=False, add_generation_prompt=True)
inputs = tok(text, return_tensors="pt").to(model.device)
out = model.generate(**inputs, max_new_tokens=8192, temperature=0.6, top_p=0.95, do_sample=True)
print(tok.decode(out[0][inputs.input_ids.shape[1]:], skip_special_tokens=True))

Full training and eval code: github.com/AdarshSingh7647/TabRanker.

Model details

• Base model: Qwen3-8B
• Method: LoRA rank 16, merged into base weights
• Precision: bfloat16, ~16 GB
• Training data: NQ Tables + MultiTabQA

Citation

@misc{singh2026tabrank,
      title={TabRank: Chain-of-Thought Distillation for Table Re-Rankers},
      author={Adarsh Singh and Kushal Raj Bhandari and Jianxi Gao and Soham Dan and Vivek Gupta},
      year={2026},
      eprint={2607.25182},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2607.25182}
}

The MultiTabQA data in this checkpoint's training mix comes from RAG over Tables:

@misc{zou2025ragtableshierarchicalmemory,
      title={RAG over Tables: Hierarchical Memory Index, Multi-Stage Retrieval, and Benchmarking},
      author={Jiaru Zou and Dongqi Fu and Sirui Chen and Xinrui He and Zihao Li and Yada Zhu and Jiawei Han and Jingrui He},
      year={2025},
      eprint={2504.01346},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2504.01346}
}