From 6660aef2451b00831373b6cc3b05818ad514a419 Mon Sep 17 00:00:00 2001 From: ModelHub XC Date: Fri, 12 Jun 2026 16:06:16 +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-Llama-3.2-1B-Full-FT-CE Source: Original Platform --- .gitattributes | 36 ++++++++++ README.md | 147 +++++++++++++++++++++++++++++++++++++++++ chat_template.jinja | 93 ++++++++++++++++++++++++++ config.json | 36 ++++++++++ generation_config.json | 14 ++++ model.safetensors | 3 + tokenizer.json | 3 + tokenizer_config.json | 15 +++++ training_args.bin | 3 + training_config.yaml | 38 +++++++++++ 10 files changed, 388 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..6831d11 --- /dev/null +++ b/README.md @@ -0,0 +1,147 @@ +--- +license: llama3.2 +base_model: meta-llama/Llama-3.2-1B-Instruct +tags: + - dia-guard + - shield + - safety + - dialect + - full-ft + - ce +language: + - en +library_name: transformers +pipeline_tag: text-generation +--- + +# Llama-3.2-1B — 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** | [`meta-llama/Llama-3.2-1B-Instruct`](https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct) | +| **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** | Llama 3.2 Community License (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-Llama-3.2-1B-Full-FT-CE", torch_dtype="bfloat16") +tokenizer = AutoTokenizer.from_pretrained("jsl5710/Shield-Llama-3.2-1B-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** | 0.71/3 (early-stopped) | +| **Train loss** | 0.5147 | +| **Train accuracy** | — | +| **Eval loss** | 0.6634 | +| **Eval accuracy** | **85.67%** | +| **Batch size (per_device × grad_accum)** | 96 × 1 = 96 | +| **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** | **0.9644** | +| **Macro Precision** | 0.9636 | +| **Macro Recall** | 0.9668 | +| **Macro F1** | **0.9642** | +| **Support** | 181,874 | + +### Per-class + +| Class | Precision | Recall | F1 | Support | +|-------|-----------|--------|----|---------| +| **safe** | 0.9311 | 0.9956 | 0.9623 | 83,140 | +| **unsafe** | 0.9961 | 0.9380 | 0.9662 | 98,734 | + +### Confusion Matrix + +| | Pred safe | Pred unsafe | +|-------------|-----------|-------------| +| **True safe** | 82,778 | 362 | +| **True unsafe** | 6,121 | 92,613 | + +> 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 **Llama 3.2 Community License**, 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..1bad6a0 --- /dev/null +++ b/chat_template.jinja @@ -0,0 +1,93 @@ +{{- bos_token }} +{%- if custom_tools is defined %} + {%- set tools = custom_tools %} +{%- endif %} +{%- if not tools_in_user_message is defined %} + {%- set tools_in_user_message = true %} +{%- endif %} +{%- if not date_string is defined %} + {%- if strftime_now is defined %} + {%- set date_string = strftime_now("%d %b %Y") %} + {%- else %} + {%- set date_string = "26 Jul 2024" %} + {%- endif %} +{%- endif %} +{%- if not tools is defined %} + {%- set tools = none %} +{%- endif %} + +{#- This block extracts the system message, so we can slot it into the right place. #} +{%- if messages[0]['role'] == 'system' %} + {%- set system_message = messages[0]['content']|trim %} + {%- set messages = messages[1:] %} +{%- else %} + {%- set system_message = "" %} +{%- endif %} + +{#- System message #} +{{- "<|start_header_id|>system<|end_header_id|>\n\n" }} +{%- if tools is not none %} + {{- "Environment: ipython\n" }} +{%- endif %} +{{- "Cutting Knowledge Date: December 2023\n" }} +{{- "Today Date: " + date_string + "\n\n" }} +{%- if tools is not none and not tools_in_user_message %} + {{- "You have access to the following functions. To call a function, please respond with JSON for a function call." }} + {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }} + {{- "Do not use variables.\n\n" }} + {%- for t in tools %} + {{- t | tojson(indent=4) }} + {{- "\n\n" }} + {%- endfor %} +{%- endif %} +{{- system_message }} +{{- "<|eot_id|>" }} + +{#- Custom tools are passed in a user message with some extra guidance #} +{%- if tools_in_user_message and not tools is none %} + {#- Extract the first user message so we can plug it in here #} + {%- if messages | length != 0 %} + {%- set first_user_message = messages[0]['content']|trim %} + {%- set messages = messages[1:] %} + {%- else %} + {{- raise_exception("Cannot put tools in the first user message when there's no first user message!") }} +{%- endif %} + {{- '<|start_header_id|>user<|end_header_id|>\n\n' -}} + {{- "Given the following functions, please respond with a JSON for a function call " }} + {{- "with its proper arguments that best answers the given prompt.\n\n" }} + {{- 'Respond in the format {"name": function name, "parameters": dictionary of argument name and its value}.' }} + {{- "Do not use variables.\n\n" }} + {%- for t in tools %} + {{- t | tojson(indent=4) }} + {{- "\n\n" }} + {%- endfor %} + {{- first_user_message + "<|eot_id|>"}} +{%- endif %} + +{%- for message in messages %} + {%- if not (message.role == 'ipython' or message.role == 'tool' or 'tool_calls' in message) %} + {{- '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' }} + {%- elif 'tool_calls' in message %} + {%- if not message.tool_calls|length == 1 %} + {{- raise_exception("This model only supports single tool-calls at once!") }} + {%- endif %} + {%- set tool_call = message.tool_calls[0].function %} + {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' -}} + {{- '{"name": "' + tool_call.name + '", ' }} + {{- '"parameters": ' }} + {{- tool_call.arguments | tojson }} + {{- "}" }} + {{- "<|eot_id|>" }} + {%- elif message.role == "tool" or message.role == "ipython" %} + {{- "<|start_header_id|>ipython<|end_header_id|>\n\n" }} + {%- if message.content is mapping or message.content is iterable %} + {{- message.content | tojson }} + {%- else %} + {{- message.content }} + {%- endif %} + {{- "<|eot_id|>" }} + {%- endif %} +{%- endfor %} +{%- if add_generation_prompt %} + {{- '<|start_header_id|>assistant<|end_header_id|>\n\n' }} +{%- endif %} diff --git a/config.json b/config.json new file mode 100644 index 0000000..758d80f --- /dev/null +++ b/config.json @@ -0,0 +1,36 @@ +{ + "architectures": [ + "LlamaForCausalLM" + ], + "attention_bias": false, + "attention_dropout": 0.0, + "bos_token_id": 128000, + "dtype": "bfloat16", + "eos_token_id": 128009, + "head_dim": 64, + "hidden_act": "silu", + "hidden_size": 2048, + "initializer_range": 0.02, + "intermediate_size": 8192, + "max_position_embeddings": 131072, + "mlp_bias": false, + "model_type": "llama", + "num_attention_heads": 32, + "num_hidden_layers": 16, + "num_key_value_heads": 8, + "pad_token_id": 128009, + "pretraining_tp": 1, + "rms_norm_eps": 1e-05, + "rope_parameters": { + "factor": 32.0, + "high_freq_factor": 4.0, + "low_freq_factor": 1.0, + "original_max_position_embeddings": 8192, + "rope_theta": 500000.0, + "rope_type": "llama3" + }, + "tie_word_embeddings": true, + "transformers_version": "5.5.0", + "use_cache": false, + "vocab_size": 128256 +} diff --git a/generation_config.json b/generation_config.json new file mode 100644 index 0000000..b88c287 --- /dev/null +++ b/generation_config.json @@ -0,0 +1,14 @@ +{ + "bos_token_id": 128000, + "do_sample": true, + "eos_token_id": [ + 128009, + 128001, + 128008, + 128009 + ], + "pad_token_id": 128009, + "temperature": 0.6, + "top_p": 0.9, + "transformers_version": "5.5.0" +} diff --git a/model.safetensors b/model.safetensors new file mode 100644 index 0000000..d3ad08c --- /dev/null +++ b/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:16e63ab68c1cbdd30977390d71ef18794e18fffb695eb36a8820e0b36b35166f +size 2471645608 diff --git a/tokenizer.json b/tokenizer.json new file mode 100644 index 0000000..1c1d8d5 --- /dev/null +++ b/tokenizer.json @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6b9e4e7fb171f92fd137b777cc2714bf87d11576700a1dcd7a399e7bbe39537b +size 17209920 diff --git a/tokenizer_config.json b/tokenizer_config.json new file mode 100644 index 0000000..f15268a --- /dev/null +++ b/tokenizer_config.json @@ -0,0 +1,15 @@ +{ + "backend": "tokenizers", + "bos_token": "<|begin_of_text|>", + "clean_up_tokenization_spaces": true, + "eos_token": "<|eot_id|>", + "is_local": false, + "model_input_names": [ + "input_ids", + "attention_mask" + ], + "model_max_length": 131072, + "pad_token": "<|eot_id|>", + "padding_side": "right", + "tokenizer_class": "TokenizersBackend" +} diff --git a/training_args.bin b/training_args.bin new file mode 100644 index 0000000..01c582b --- /dev/null +++ b/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:05d22369a0e6dd627b1f730b89fcbc9dc467e6e13a957fd6e09a968e25fe7098 +size 5777 diff --git a/training_config.yaml b/training_config.yaml new file mode 100644 index 0000000..cb9b582 --- /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: 3.0e-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: meta-llama/Llama-3.2-1B-Instruct +num_epochs: 3 +output_dir: /data/vibe_exp/dia-guard/models/group3_student_ft_baseline/full_ft/llama_3_2_1b_instruct +per_device_eval_batch_size: 96 +per_device_train_batch_size: 96 +report_to: wandb +run_name: llama-3.2-1b-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