Add files using upload-large-folder tool

.gitattributes

@@ -1,35 +1 @@
-*.7z filter=lfs diff=lfs merge=lfs -text
-*.arrow filter=lfs diff=lfs merge=lfs -text
-*.bin filter=lfs diff=lfs merge=lfs -text
-*.bz2 filter=lfs diff=lfs merge=lfs -text
-*.ckpt filter=lfs diff=lfs merge=lfs -text
-*.ftz filter=lfs diff=lfs merge=lfs -text
-*.gz filter=lfs diff=lfs merge=lfs -text
-*.h5 filter=lfs diff=lfs merge=lfs -text
-*.joblib filter=lfs diff=lfs merge=lfs -text
-*.lfs.* filter=lfs diff=lfs merge=lfs -text
-*.mlmodel filter=lfs diff=lfs merge=lfs -text
-*.model filter=lfs diff=lfs merge=lfs -text
-*.msgpack filter=lfs diff=lfs merge=lfs -text
-*.npy filter=lfs diff=lfs merge=lfs -text
-*.npz filter=lfs diff=lfs merge=lfs -text
-*.onnx filter=lfs diff=lfs merge=lfs -text
-*.ot filter=lfs diff=lfs merge=lfs -text
-*.parquet filter=lfs diff=lfs merge=lfs -text
-*.pb filter=lfs diff=lfs merge=lfs -text
-*.pickle filter=lfs diff=lfs merge=lfs -text
-*.pkl filter=lfs diff=lfs merge=lfs -text
-*.pt filter=lfs diff=lfs merge=lfs -text
-*.pth filter=lfs diff=lfs merge=lfs -text
-*.rar filter=lfs diff=lfs merge=lfs -text
 *.safetensors filter=lfs diff=lfs merge=lfs -text
-saved_model/**/* filter=lfs diff=lfs merge=lfs -text
-*.tar.* filter=lfs diff=lfs merge=lfs -text
-*.tar filter=lfs diff=lfs merge=lfs -text
-*.tflite filter=lfs diff=lfs merge=lfs -text
-*.tgz filter=lfs diff=lfs merge=lfs -text
-*.wasm filter=lfs diff=lfs merge=lfs -text
-*.xz filter=lfs diff=lfs merge=lfs -text
-*.zip filter=lfs diff=lfs merge=lfs -text
-*.zst filter=lfs diff=lfs merge=lfs -text
-*tfevents* filter=lfs diff=lfs merge=lfs -text

README.md

@@ -0,0 +1,174 @@
+---
+base_model: google/gemma-3-4b-it
+library_name: peft
+pipeline_tag: text-generation
+language:
+- en
+tags:
+- peft
+- lora
+- data-to-text
+- text-to-data
+- factual-consistency
+- hallucination-detection
+---
+
+# XQDT E2E verifier: gemma3 4B
+
+This repository contains the LoRA adapter for the **gemma3 4B**
+XQDT verifier reported in *XQDT: eXplainable and Quantitative Data-Text Alignment
+Metric with Feedback Signals*. The base model is
+[`google/gemma-3-4b-it`](https://huggingface.co/google/gemma-3-4b-it); base-model
+weights are not included here.
+
+This checkpoint corresponds to the paper's E2E setting and was trained on the joint WebNLG--E2E synthetic training set; the short repository name does not mean E2E-only training.
+
+## Intended use
+
+XQDT verifies alignment between English text and structured triples. It returns
+`missing`, `extra`, and `incorrect` units, or `All correct`. In the paper's
+terminology, `missing` identifies an input unit omitted from the text; `extra`
+identifies text content unsupported by the input; and `incorrect` identifies an
+input unit realised with incorrect information.
+
+The expected prompt and four frozen regression examples are provided in
+`smoke_test.json`. This model is an evaluation component, not a general-purpose
+fact checker, and has not been validated outside data--text alignment settings.
+
+## Prompt format
+
+```text
+Verify if the triples align with the text. Find missing, extra, or incorrect triples.
+TEXT: {text}
+TRIPLES:
+1. [S] {subject} [P] {predicate} [O] {object}
+Output as markdown table with Type and Triple columns.
+```
+
+The canonical paper implementation uses **ms-swift PtEngine**. Transformers and
+vLLM are provided as portable alternatives. Different libraries, versions, and
+sampling implementations can produce small output differences; compare parsed
+error units rather than requiring byte-identical text.
+
+## ms-swift PtEngine
+
+```python
+import torch
+from huggingface_hub import snapshot_download
+from swift.llm import InferRequest, PtEngine, RequestConfig, get_model_tokenizer, get_template
+from swift.tuners import Swift
+
+BASE_MODEL = "google/gemma-3-4b-it"
+ADAPTER_ID = "Loria-MosAIk/xqdt-e2e-gemma3-4b"
+SYSTEM_PROMPT = "Identify extra, missing and incorrect triples precisely."
+QUERY = """Verify if the triples align with the text. Find missing, extra, or incorrect triples.
+TEXT: Blue Spice is a coffee shop in city centre.
+TRIPLES:
+1. [S] Blue Spice [P] area [O] city centre
+2. [S] Blue Spice [P] eat type [O] coffee shop
+Output as markdown table with Type and Triple columns."""
+
+adapter_path = snapshot_download(ADAPTER_ID)
+model, tokenizer = get_model_tokenizer(
+    BASE_MODEL,
+    model_kwargs={"device_map": "auto", "torch_dtype": torch.bfloat16},
+)
+model = Swift.from_pretrained(model, model_id=adapter_path, adapter_name="default")
+template = get_template("gemma3_text", tokenizer, default_system=None)
+engine = PtEngine.from_model_template(model, template, max_batch_size=1)
+response = engine.infer(
+    [InferRequest(messages=[{"role": "user", "content": QUERY}])],
+    RequestConfig(max_tokens=1024, temperature=0.3),
+    use_tqdm=False,
+)[0]
+print(response.choices[0].message.content)
+```
+
+## Transformers and PEFT
+
+```python
+import torch
+from peft import PeftModel
+from transformers import set_seed
+
+BASE_MODEL = "google/gemma-3-4b-it"
+ADAPTER_ID = "Loria-MosAIk/xqdt-e2e-gemma3-4b"
+SYSTEM_PROMPT = "Identify extra, missing and incorrect triples precisely."
+QUERY = """Verify if the triples align with the text. Find missing, extra, or incorrect triples.
+TEXT: Blue Spice is a coffee shop in city centre.
+TRIPLES:
+1. [S] Blue Spice [P] area [O] city centre
+2. [S] Blue Spice [P] eat type [O] coffee shop
+Output as markdown table with Type and Triple columns."""
+MESSAGES = [{"role": "user", "content": QUERY}]
+set_seed(2023)
+
+from transformers import AutoProcessor, Gemma3ForConditionalGeneration
+
+processor = AutoProcessor.from_pretrained(BASE_MODEL)
+base = Gemma3ForConditionalGeneration.from_pretrained(
+    BASE_MODEL, torch_dtype=torch.bfloat16, device_map="auto"
+)
+model = PeftModel.from_pretrained(base, ADAPTER_ID).eval()
+prompt = processor.apply_chat_template(MESSAGES, tokenize=False, add_generation_prompt=True)
+inputs = processor(text=prompt, return_tensors="pt").to(model.device)
+with torch.inference_mode():
+    output = model.generate(**inputs, max_new_tokens=1024, do_sample=True, temperature=0.3)
+generated = output[0, inputs["input_ids"].shape[-1]:]
+print(processor.decode(generated, skip_special_tokens=True))
+```
+
+## vLLM
+
+The reference vLLM environment uses NVIDIA H100 hardware and the pinned versions
+listed in `requirements.txt`. Other recent CUDA GPUs may also work, but are treated
+as best-effort environments and should be recorded in the smoke-test report.
+
+```python
+from huggingface_hub import snapshot_download
+from vllm import LLM, SamplingParams
+from vllm.lora.request import LoRARequest
+
+BASE_MODEL = "google/gemma-3-4b-it"
+ADAPTER_ID = "Loria-MosAIk/xqdt-e2e-gemma3-4b"
+SYSTEM_PROMPT = "Identify extra, missing and incorrect triples precisely."
+QUERY = """Verify if the triples align with the text. Find missing, extra, or incorrect triples.
+TEXT: Blue Spice is a coffee shop in city centre.
+TRIPLES:
+1. [S] Blue Spice [P] area [O] city centre
+2. [S] Blue Spice [P] eat type [O] coffee shop
+Output as markdown table with Type and Triple columns."""
+MESSAGES = [{"role": "user", "content": QUERY}]
+
+adapter_path = snapshot_download(ADAPTER_ID)
+llm = LLM(model=BASE_MODEL, enable_lora=True)
+outputs = llm.chat(
+    MESSAGES,
+    SamplingParams(max_tokens=1024, temperature=0.3, seed=2023),
+    lora_request=LoRARequest("xqdt", 1, adapter_path),
+)
+print(outputs[0].outputs[0].text)
+```
+
+## Reproducibility
+
+- Adapter SHA-256 and sanitized training hyperparameters: `training_manifest.json`
+- Frozen inputs and reference outputs: `smoke_test.json`
+- Paper: [https://openreview.net/forum?id=t1037gQHuf](https://openreview.net/forum?id=t1037gQHuf)
+- Code: [https://github.com/guihuzhang/xqdt](https://github.com/guihuzhang/xqdt)
+
+The original training run did not freeze a public Hugging Face revision for every
+base model. The standard base-model ID above replaces the machine-local cache path
+stored by the training framework. Users must comply with the corresponding base
+model's access terms and license.
+
+## Citation
+
+```bibtex
+@inproceedings{efimov-zhang-etal-2026-xqdt,
+  title = {XQDT: eXplainable and Quantitative Data-Text Alignment Metric with Feedback Signals},
+  author = {Efimov-Zhang, Kun and Song, Yifei and Gardent, Claire},
+  booktitle = {Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing},
+  year = {2026}
+}
+```

adapter_config.json

@@ -0,0 +1,38 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "google/gemma-3-4b-it",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 32,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": [],
+  "peft_type": "LORA",
+  "peft_version": "0.18.0",
+  "qalora_group_size": 16,
+  "r": 16,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": "^(model.language_model.*\\.(k_proj|down_proj|o_proj|q_proj|gate_proj|v_proj|up_proj))$",
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:410f60027bb0884923650616b40eeeca780063887717817e8444f91e218775b8
+size 119280712

additional_config.json

@@ -0,0 +1 @@
+{"lora_dtype": null, "lorap_lr_ratio": null, "lorap_emb_lr": 1e-06}

provenance.json

@@ -0,0 +1,15 @@
+{
+  "original_base_model_reference_sha256": "f86883a98bbc65b60c2e2f8659ceb082787492caa846bfc930d8bc798b2f0bbf",
+  "original_base_model_reference_was_machine_local": true,
+  "published_base_model_name_or_path": "google/gemma-3-4b-it",
+  "schema_version": 1,
+  "source_checkpoint_relative_path": "checkpoints/checkpoints_e2e19/output_gemma3_4b_e2e19/v0-20260215-222239/checkpoint-6975",
+  "source_commit": "739c84d868f9d32bcf5c150bdf05c56260cfcaf4",
+  "source_files": {
+    "adapter_config.json": "15370905d97b26cc099b3733a8a1fbd7c00078357eb275cd5aeb0404cc3da037",
+    "adapter_model.safetensors": "410f60027bb0884923650616b40eeeca780063887717817e8444f91e218775b8",
+    "additional_config.json": "c7799462ebedae6557ffad31566029e2d2f958b7b40e46e972cf901bcaf45733",
+    "args.json": "f017bae22f7b4bc446dfe1afdf769656df7cb823b3e1ad386e6230c1f5f5f601",
+    "reference_predictions": "51b9c4e077f7495ca02acbeca9d5798bdbee1ea498c9ec60bf13fd98c08886e0"
+  }
+}

requirements.txt

@@ -0,0 +1,7 @@
+torch==2.10.0
+ms-swift==4.1.3
+transformers==4.57.6
+peft==0.19.1
+huggingface-hub==0.36.2
+safetensors==0.7.0
+vllm==0.18.1

smoke_test.json

@@ -0,0 +1,159 @@
+{
+  "base_model": "google/gemma-3-4b-it",
+  "cases": [
+    {
+      "case": "correct",
+      "gold": {
+        "extra": [],
+        "incorrect": [],
+        "label": "positive",
+        "missing": []
+      },
+      "input_text": "Blue Spice is a coffee shop in city centre.",
+      "input_triples": [
+        "Blue Spice | area | city centre",
+        "Blue Spice | eat type | coffee shop"
+      ],
+      "reference_model_output": "All correct",
+      "reference_parsed_output": {
+        "extra": [],
+        "incorrect": [],
+        "missing": []
+      },
+      "sample_id": 0,
+      "sample_index": 0
+    },
+    {
+      "case": "omitted",
+      "gold": {
+        "extra": [],
+        "incorrect": [],
+        "label": "negative",
+        "missing": [
+          "Mango | buenos aires | Crowne Plaza Hotel",
+          "The Cricketers | family friendly | yes",
+          "The Cricketers | eat type | restaurant"
+        ]
+      },
+      "input_text": "The coffee shop Blue Spice is based near Crowne Plaza Hotel and has a high customer rating of 5 out of 5.",
+      "input_triples": [
+        "Blue Spice | customer rating | 5 out of 5",
+        "Blue Spice | eat type | coffee shop",
+        "Blue Spice | near | Crowne Plaza Hotel",
+        "Mango | buenos aires | Crowne Plaza Hotel",
+        "The Cricketers | family friendly | yes",
+        "The Cricketers | eat type | restaurant"
+      ],
+      "reference_model_output": "| Type | Triple |\n| ---- | ------ |\n| Missing | [S] Mango [P] buenos aires [O] Crowne Plaza Hotel |\n| Missing | [S] The Cricketers [P] eat type [O] restaurant |\n| Missing | [S] The Cricketers [P] family friendly [O] yes |",
+      "reference_parsed_output": {
+        "extra": [],
+        "incorrect": [],
+        "missing": [
+          "[S] Mango [P] buenos aires [O] Crowne Plaza Hotel",
+          "[S] The Cricketers [P] eat type [O] restaurant",
+          "[S] The Cricketers [P] family friendly [O] yes"
+        ]
+      },
+      "sample_id": 2,
+      "sample_index": 2
+    },
+    {
+      "case": "extra",
+      "gold": {
+        "extra": [
+          "Blue Spice | eat type | pub"
+        ],
+        "incorrect": [],
+        "label": "negative",
+        "missing": []
+      },
+      "input_text": "At the riverside, there is a pub called The Blue Spice.",
+      "input_triples": [
+        "Blue Spice | area | riverside"
+      ],
+      "reference_model_output": "| Type | Triple |\n| ---- | ------ |\n| Extra | [S] Blue Spice [P] eat type [O] pub |",
+      "reference_parsed_output": {
+        "extra": [
+          "[S] Blue Spice [P] eat type [O] pub"
+        ],
+        "incorrect": [],
+        "missing": []
+      },
+      "sample_id": 6,
+      "sample_index": 6
+    },
+    {
+      "case": "incorrect",
+      "gold": {
+        "extra": [],
+        "incorrect": [
+          {
+            "error_type": "wrong_entity",
+            "incorrect": "Hot | area | riverside",
+            "original": "Blue Spice | area | riverside",
+            "replaced_element": "subject"
+          },
+          {
+            "error_type": "wrong_entity",
+            "incorrect": "Blue Spice | family friendly | Politics",
+            "original": "Blue Spice | family friendly | no",
+            "replaced_element": "object"
+          },
+          {
+            "error_type": "wrong_predicate",
+            "incorrect": "Blue Spice | brand | pub",
+            "original": "Blue Spice | eat type | pub",
+            "replaced_element": "predicate"
+          },
+          {
+            "error_type": "wrong_predicate",
+            "incorrect": "Blue Spice | address | Rainbow Vegetarian Café",
+            "original": "Blue Spice | near | Rainbow Vegetarian Café",
+            "replaced_element": "predicate"
+          },
+          {
+            "error_type": "wrong_predicate",
+            "incorrect": "Blue Spice | class | Chinese",
+            "original": "Blue Spice | food | Chinese",
+            "replaced_element": "predicate"
+          }
+        ],
+        "label": "negative",
+        "missing": []
+      },
+      "input_text": "Blue Spice pub in riverside serves Chinese food. It is not family friendly and can be found near Rainbow Vegetarian Café.",
+      "input_triples": [
+        "Hot | area | riverside",
+        "Blue Spice | brand | pub",
+        "Blue Spice | family friendly | Politics",
+        "Blue Spice | class | Chinese",
+        "Blue Spice | address | Rainbow Vegetarian Café"
+      ],
+      "reference_model_output": "| Type | Triple |\n| ---- | ------ |\n| Incorrect | [S] Hot [P] area [O] riverside |\n| Incorrect | [S] Blue Spice [P] brand [O] pub |\n| Incorrect | [S] Blue Spice [P] family friendly [O] Politics |\n| Incorrect | [S] Blue Spice [P] class [O] Chinese |\n| Incorrect | [S] Blue Spice [P] address [O] Rainbow Vegetarian Café |",
+      "reference_parsed_output": {
+        "extra": [],
+        "incorrect": [
+          "[S] Hot [P] area [O] riverside",
+          "[S] Blue Spice [P] brand [O] pub",
+          "[S] Blue Spice [P] family friendly [O] Politics",
+          "[S] Blue Spice [P] class [O] Chinese",
+          "[S] Blue Spice [P] address [O] Rainbow Vegetarian Café"
+        ],
+        "missing": []
+      },
+      "sample_id": 13,
+      "sample_index": 13
+    }
+  ],
+  "comparison": "Compare parsed error units after normalization. Byte-identical generation is not required across inference libraries.",
+  "reference_backend": "ms-swift PtEngine",
+  "reference_generation": {
+    "max_tokens": 1024,
+    "seed": 2023,
+    "temperature": 0.3
+  },
+  "reference_predictions_relative_path": "verifier_train_eval_e2e19/xqdt_outputs/test_predictions_gemma3_4b_6975.json",
+  "reference_predictions_sha256": "51b9c4e077f7495ca02acbeca9d5798bdbee1ea498c9ec60bf13fd98c08886e0",
+  "repository": "xqdt-e2e-gemma3-4b",
+  "schema_version": 1
+}

training_manifest.json

@@ -0,0 +1,45 @@
+{
+  "adapter_model_sha256": "410f60027bb0884923650616b40eeeca780063887717817e8444f91e218775b8",
+  "base_model": "google/gemma-3-4b-it",
+  "dataset_release_label": "e2e",
+  "hyperparameters": {
+    "bf16": true,
+    "dataset": [
+      "triple_error_train"
+    ],
+    "dataset_num_proc": 1,
+    "eval_steps": 279,
+    "fp16": false,
+    "gradient_accumulation_steps": 4,
+    "gradient_checkpointing": true,
+    "learning_rate": 5e-05,
+    "logging_steps": 93,
+    "lora_alpha": 32,
+    "lora_dropout": 0.05,
+    "lora_rank": 16,
+    "max_length": 1024,
+    "max_new_tokens": 64,
+    "model": "google/gemma-3-4b-it",
+    "model_type": "gemma3_vision",
+    "num_train_epochs": 16,
+    "per_device_eval_batch_size": 4,
+    "per_device_train_batch_size": 4,
+    "save_steps": 279,
+    "save_total_limit": 3,
+    "seed": 2023,
+    "system": null,
+    "target_modules": [
+      "all-linear"
+    ],
+    "template": "gemma3_text",
+    "torch_dtype": "bfloat16",
+    "train_type": "lora",
+    "warmup_ratio": 0.0625,
+    "weight_decay": 0.01
+  },
+  "repository": "xqdt-e2e-gemma3-4b",
+  "schema_version": 1,
+  "source_checkpoint_relative_path": "checkpoints/checkpoints_e2e19/output_gemma3_4b_e2e19/v0-20260215-222239/checkpoint-6975",
+  "source_commit": "739c84d868f9d32bcf5c150bdf05c56260cfcaf4",
+  "training_scope": "joint WebNLG and E2E synthetic data"
+}