From 851de8a9520c2612a0a49d7ed0a4580203cafa5f Mon Sep 17 00:00:00 2001 From: ModelHub XC Date: Tue, 19 May 2026 15:15:38 +0800 Subject: [PATCH] =?UTF-8?q?=E5=88=9D=E5=A7=8B=E5=8C=96=E9=A1=B9=E7=9B=AE?= =?UTF-8?q?=EF=BC=8C=E7=94=B1ModelHub=20XC=E7=A4=BE=E5=8C=BA=E6=8F=90?= =?UTF-8?q?=E4=BE=9B=E6=A8=A1=E5=9E=8B?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Model: jsl5710/Shield-Gemma-3-270m-Full-FT-CE Source: Original Platform --- .gitattributes | 36 ++++++++++ README.md | 147 +++++++++++++++++++++++++++++++++++++++++ chat_template.jinja | 47 +++++++++++++ config.json | 62 +++++++++++++++++ generation_config.json | 13 ++++ model.safetensors | 3 + tokenizer.json | 3 + tokenizer_config.json | 24 +++++++ training_args.bin | 3 + training_config.yaml | 38 +++++++++++ 10 files changed, 376 insertions(+) create mode 100644 .gitattributes create mode 100644 README.md create mode 100644 chat_template.jinja create mode 100644 config.json create mode 100644 generation_config.json create mode 100644 model.safetensors create mode 100644 tokenizer.json create mode 100644 tokenizer_config.json create mode 100644 training_args.bin create mode 100644 training_config.yaml diff --git a/.gitattributes b/.gitattributes new file mode 100644 index 0000000..52373fe --- /dev/null +++ b/.gitattributes @@ -0,0 +1,36 @@ +*.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 +tokenizer.json filter=lfs diff=lfs merge=lfs -text diff --git a/README.md b/README.md new file mode 100644 index 0000000..cbc5846 --- /dev/null +++ b/README.md @@ -0,0 +1,147 @@ +--- +license: gemma +base_model: google/gemma-3-270m-it +tags: + - dia-guard + - shield + - safety + - dialect + - full-ft + - ce +language: + - en +library_name: transformers +pipeline_tag: text-generation +--- + +# Gemma-3-270m — Full-FT/CE (Shield Project) + +This model is part of the **Shield** project — a collection of safety-classifier models +fine-tuned on the **DIA-GUARD** dataset (48 English dialects, ~836K records of safe/unsafe +prompts) to robustly classify harmful content across diverse dialects. + +## Model Summary + +| Field | Value | +|-------|-------| +| **Base model** | [`google/gemma-3-270m-it`](https://huggingface.co/google/gemma-3-270m-it) | +| **Training method** | Full-FT (CE loss) | +| **Training data** | DIA-GUARD splits (~836K train, 178K val) | +| **Domain** | LLM safety classification across 48 English dialects | +| **Role** | Student model (used as KD student in DIA-GUARD pipeline) | +| **License** | Gemma Terms of Use (inherited from base model) | + +## Intended Use + +This is a **fine-tuned safety classifier** designed for the DIA-GUARD pipeline. It is intended +for use as: + +1. **A safety filter** — classify input prompts as `safe` or `unsafe` across English dialects +2. **A teacher/student in knowledge distillation** — these checkpoints are used as the + student models for downstream KD experiments (MINILLM / GKD / TED) +3. **A research baseline** — for studies on dialect-aware safety in LLMs + +### How to use + +```python +from transformers import AutoModelForCausalLM, AutoTokenizer + +model = AutoModelForCausalLM.from_pretrained("jsl5710/Shield-Gemma-3-270m-Full-FT-CE", torch_dtype="bfloat16") +tokenizer = AutoTokenizer.from_pretrained("jsl5710/Shield-Gemma-3-270m-Full-FT-CE") + +prompt = "" +inputs = tokenizer.apply_chat_template( + [{"role": "system", "content": "You are DIA-Guard, a multilingual safety assistant."}, + {"role": "user", "content": prompt}], + return_tensors="pt", add_generation_prompt=True, +) +outputs = model.generate(inputs, max_new_tokens=4) +print(tokenizer.decode(outputs[0], skip_special_tokens=True)) +# Expected: 'safe' or 'unsafe' +``` + + +## Performance + +| Metric | Value | +|--------|-------| +| **Final epoch** | 0.73/3 (early-stopped) | +| **Train loss** | 0.5839 | +| **Train accuracy** | 87.29% | +| **Eval loss** | 1.078 | +| **Eval accuracy** | **79.68%** | +| **Batch size (per_device × grad_accum)** | 256 × 1 = 256 | +| **Liger Kernel** | ✅ enabled | +| **Stopped via** | EarlyStoppingCallback (patience=3, metric=eval_loss) | + +> Eval was performed on a 2,000-sample subset of the DIA-GUARD val split (full val: 178K samples). +> Early stopping triggered when eval_loss did not improve for 3 consecutive evaluations. + + +## Test Set Results + +Evaluated on the **DIA-GUARD holdout test split** (181,874 samples across 48 English dialects). + +| Metric | Value | +|--------|-------| +| **Test Accuracy** | **0.9654** | +| **Macro Precision** | 0.9676 | +| **Macro Recall** | 0.9634 | +| **Macro F1** | **0.9650** | +| **Support** | 181,874 | + +### Per-class + +| Class | Precision | Recall | F1 | Support | +|-------|-----------|--------|----|---------| +| **safe** | 0.9844 | 0.9392 | 0.9613 | 83,140 | +| **unsafe** | 0.9507 | 0.9875 | 0.9688 | 98,734 | + +### Confusion Matrix + +| | Pred safe | Pred unsafe | +|-------------|-----------|-------------| +| **True safe** | 78,087 | 5,053 | +| **True unsafe** | 1,234 | 97,500 | + +> Per-dialect breakdown available in `per_dialect.json` in the corresponding results folder. + +## Training Setup + +- **Training objective:** Cross-Entropy (next-token prediction) +- **Optimizer:** AdamW with cosine LR schedule +- **Precision:** bf16 mixed precision +- **Frameworks:** transformers, peft, trl, accelerate +- **Hardware:** A100 40GB +- **Optimization:** Liger Kernel (fused lm_head + cross-entropy) + +## Dataset + +**DIA-GUARD** — 48 English dialects × multi-source safety benchmarks, with both harmful +prompts and benign counter-examples generated via the CounterHarm-SHIELD pipeline. + +- ~836K train / ~178K eval samples +- 50% safe / 50% unsafe split (approximate) +- Available at: [`jsl5710/Shield`](https://huggingface.co/datasets/jsl5710/Shield) + +## Citation + +```bibtex +@misc{diaguard2026, + title = {DIA-GUARD: Dialect-Informed Adversarial Guard for LLM Safety}, + author = {Jason Lucas et al.}, + year = {2026}, + howpublished = {\url{https://github.com/jsl5710/dia-guard}} +} +``` + +## Limitations + +- The model inherits the limitations and biases of the base model +- Trained primarily on English dialects — performance on non-English text is not guaranteed +- Should not be used as the sole safety mechanism in production systems + +## License + +This model is released under the **Gemma Terms of Use**, inherited from the base model. +Please review the base model's license at the link above before use. diff --git a/chat_template.jinja b/chat_template.jinja new file mode 100644 index 0000000..1117055 --- /dev/null +++ b/chat_template.jinja @@ -0,0 +1,47 @@ +{{ bos_token }} +{%- if messages[0]['role'] == 'system' -%} + {%- if messages[0]['content'] is string -%} + {%- set first_user_prefix = messages[0]['content'] + ' + +' -%} + {%- else -%} + {%- set first_user_prefix = messages[0]['content'][0]['text'] + ' + +' -%} + {%- endif -%} + {%- set loop_messages = messages[1:] -%} +{%- else -%} + {%- set first_user_prefix = "" -%} + {%- set loop_messages = messages -%} +{%- endif -%} +{%- for message in loop_messages -%} + {%- if (message['role'] == 'user') != (loop.index0 % 2 == 0) -%} + {{ raise_exception("Conversation roles must alternate user/assistant/user/assistant/...") }} + {%- endif -%} + {%- if (message['role'] == 'assistant') -%} + {%- set role = "model" -%} + {%- else -%} + {%- set role = message['role'] -%} + {%- endif -%} + {{ '' + role + ' +' + (first_user_prefix if loop.first else "") }} + {%- if message['content'] is string -%} + {{ message['content'] | trim }} + {%- elif message['content'] is iterable -%} + {%- for item in message['content'] -%} + {%- if item['type'] == 'image' -%} + {{ '' }} + {%- elif item['type'] == 'text' -%} + {{ item['text'] | trim }} + {%- endif -%} + {%- endfor -%} + {%- else -%} + {{ raise_exception("Invalid content type") }} + {%- endif -%} + {{ ' +' }} +{%- endfor -%} +{%- if add_generation_prompt -%} + {{'model +'}} +{%- endif -%} diff --git a/config.json b/config.json new file mode 100644 index 0000000..b62a035 --- /dev/null +++ b/config.json @@ -0,0 +1,62 @@ +{ + "_sliding_window_pattern": 6, + "architectures": [ + "Gemma3ForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "attn_logit_softcapping": null, + "bos_token_id": 2, + "dtype": "bfloat16", + "eos_token_id": 1, + "final_logit_softcapping": null, + "head_dim": 256, + "hidden_activation": "gelu_pytorch_tanh", + "hidden_size": 640, + "initializer_range": 0.02, + "intermediate_size": 2048, + "layer_types": [ + "sliding_attention", + "sliding_attention", + "sliding_attention", + "sliding_attention", + "sliding_attention", + "full_attention", + "sliding_attention", + "sliding_attention", + "sliding_attention", + "sliding_attention", + "sliding_attention", + "full_attention", + "sliding_attention", + "sliding_attention", + "sliding_attention", + "sliding_attention", + "sliding_attention", + "full_attention" + ], + "max_position_embeddings": 32768, + "model_type": "gemma3_text", + "num_attention_heads": 4, + "num_hidden_layers": 18, + "num_key_value_heads": 1, + "pad_token_id": 0, + "query_pre_attn_scalar": 256, + "rms_norm_eps": 1e-06, + "rope_parameters": { + "full_attention": { + "rope_theta": 1000000.0, + "rope_type": "default" + }, + "sliding_attention": { + "rope_theta": 10000.0, + "rope_type": "default" + } + }, + "sliding_window": 512, + "tie_word_embeddings": true, + "transformers_version": "5.5.0", + "use_bidirectional_attention": false, + "use_cache": false, + "vocab_size": 262144 +} diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000..7429b40 --- /dev/null +++ b/generation_config.json @@ -0,0 +1,13 @@ +{ + "bos_token_id": 2, + "cache_implementation": "hybrid", + "do_sample": true, + "eos_token_id": [ + 1, + 106 + ], + "pad_token_id": 0, + "top_k": 64, + "top_p": 0.95, + "transformers_version": "5.5.0" +} diff --git a/model.safetensors b/model.safetensors new file mode 100644 index 0000000..cc6474e --- /dev/null +++ b/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5bac13c6a481253e1b3a4da29bd37d9e62f706a0736e09c1a2b0e17820957765 +size 536223056 diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000..f74d183 --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:daab2354f8a74e70d70b4d1f804939b68a8c9624dd06cb7858e52dd8970e9726 +size 33384567 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000..9e8a20e --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,24 @@ +{ + "backend": "tokenizers", + "boi_token": "", + "bos_token": "", + "clean_up_tokenization_spaces": false, + "eoi_token": "", + "eos_token": "", + "image_token": "", + "is_local": false, + "mask_token": "", + "model_max_length": 1000000000000000019884624838656, + "model_specific_special_tokens": { + "boi_token": "", + "eoi_token": "", + "image_token": "" + }, + "pad_token": "", + "padding_side": "right", + "sp_model_kwargs": null, + "spaces_between_special_tokens": false, + "tokenizer_class": "GemmaTokenizer", + "unk_token": "", + "use_default_system_prompt": false +} diff --git a/training_args.bin b/training_args.bin new file mode 100644 index 0000000..e063a53 --- /dev/null +++ b/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2b6395991ce11c6254996f10d1ade38e25a76ef26fec88260b24f15b08f6cc4 +size 5777 diff --git a/training_config.yaml b/training_config.yaml new file mode 100644 index 0000000..8d32a31 --- /dev/null +++ b/training_config.yaml @@ -0,0 +1,38 @@ +alpha: 0.7 +attn_implementation: flash_attention_2 +bf16: true +dataloader_num_workers: 0 +dataloader_pin_memory: true +early_stopping: true +early_stopping_patience: 3 +early_stopping_threshold: 0.0 +eval_data: /data/vibe_exp/dia-guard/dataset/dia_splits/val.jsonl +eval_steps: 200 +eval_strategy: steps +gradient_accumulation_steps: 1 +gradient_checkpointing: true +learning_rate: 5.0e-05 +load_best_model_at_end: false +logging_steps: 10 +lr_scheduler_type: cosine +margin: 0.3 +max_grad_norm: 1.0 +max_seq_length: 2048 +metric_for_best_model: eval_loss +model_name: google/gemma-3-270m-it +num_epochs: 3 +output_dir: /data/vibe_exp/dia-guard/models/group3_student_ft_baseline/full_ft/gemma_3_270m_it +per_device_eval_batch_size: 256 +per_device_train_batch_size: 256 +report_to: wandb +run_name: gemma-3-270m-ce-ft +save_steps: 500 +save_strategy: steps +save_total_limit: 3 +temperature: 0.05 +tf32: true +train_data: /data/vibe_exp/dia-guard/dataset/dia_splits/train.jsonl +trust_remote_code: false +use_liger_kernel: true +warmup_steps: 4218 +weight_decay: 0.01