| | --- |
| | tags: |
| | - w8a8 |
| | - vllm |
| | language: |
| | - en |
| | - zh |
| | pipeline_tag: text-generation |
| | base_model: zai-org/GLM-4.6 |
| | --- |
| | |
| | # GLM-4.6-quantized.w8a8 |
| |
|
| | ## Model Overview |
| | - **Model Architecture:** zai-org/GLM-4.6 |
| | - **Input:** Text |
| | - **Output:** Text |
| | - **Model Optimizations:** |
| | - **Weight quantization:** INT8 |
| | - **Activation quantization:** INT8 |
| | - **Out-of-scope:** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in languages other than English. |
| | - **Version:** 1.0 |
| | - **Model Developers:** RedHatAI |
| |
|
| | This model is a quantized version of [zai-org/GLM-4.6](https://huggingface.co/zai-org/GLM-4.6). |
| | It was evaluated on a several tasks to assess the its quality in comparison to the unquatized model. |
| |
|
| | ### Model Optimizations |
| |
|
| | This model was obtained by quantizing the weights and activations of [zai-org/GLM-4.6](https://huggingface.co/zai-org/GLM-4.6) to INT8 data type, ready for inference with vLLM>=0.11.0. |
| |
|
| | Only the weights and activations of the linear operators within transformers blocks are quantized using [LLM Compressor](https://github.com/vllm-project/llm-compressor). |
| |
|
| | ## Deployment |
| |
|
| | ### Use with vLLM |
| |
|
| | This model can be deployed efficiently using the [vLLM](https://docs.vllm.ai/en/latest/) backend, as shown in the example below. |
| |
|
| | ```python |
| | from vllm import LLM, SamplingParams |
| | from transformers import AutoTokenizer |
| | |
| | model_id = "RedHatAI/GLM-4.6-quantized.w8a8" |
| | number_gpus = 4 |
| | |
| | sampling_params = SamplingParams(temperature=0.6, top_p=0.9, max_tokens=256) |
| | |
| | tokenizer = AutoTokenizer.from_pretrained(model_id) |
| | |
| | messages = [ |
| | {"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"}, |
| | {"role": "user", "content": "Who are you?"}, |
| | ] |
| | |
| | prompts = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False) |
| | |
| | llm = LLM(model=model_id, tensor_parallel_size=number_gpus) |
| | |
| | outputs = llm.generate(prompts, sampling_params) |
| | |
| | generated_text = outputs[0].outputs[0].text |
| | print(generated_text) |
| | ``` |
| |
|
| | vLLM aslo supports OpenAI-compatible serving. See the [documentation](https://docs.vllm.ai/en/latest/) for more details. |
| |
|
| | ## Creation |
| |
|
| | This model was created by applying a script similar to [LLM Compressor with calibration samples from UltraChat](https://github.com/vllm-project/llm-compressor/blob/main/examples/quantizing_moe/glm4_7_example.py), as presented in the code snipet below. |
| |
|
| | <details> |
| | |
| | ```python |
| | from datasets import load_dataset |
| | from transformers import AutoModelForCausalLM, AutoTokenizer |
| | |
| | from llmcompressor import oneshot |
| | from llmcompressor.modifiers.quantization import GPTQModifier |
| | from llmcompressor.utils import dispatch_for_generation |
| | |
| | MODEL_ID = "zai-org/GLM-4.6" |
| | |
| | # Load model. |
| | model = AutoModelForCausalLM.from_pretrained( |
| | MODEL_ID, torch_dtype="auto" |
| | ) |
| | tokenizer = AutoTokenizer.from_pretrained(MODEL_ID) |
| | |
| | # Select calibration dataset. |
| | DATASET_ID = "HuggingFaceH4/ultrachat_200k" |
| | DATASET_SPLIT = "train_sft" |
| | |
| | # Select number of samples. |
| | # Increasing the number of samples can improve accuracy. |
| | NUM_CALIBRATION_SAMPLES = 512 |
| | MAX_SEQUENCE_LENGTH = 2048 |
| | |
| | # Load dataset and preprocess. |
| | ds = load_dataset(DATASET_ID, split=f"{DATASET_SPLIT}[:{NUM_CALIBRATION_SAMPLES}]") |
| | ds = ds.shuffle(seed=42) |
| | |
| | def preprocess(example): |
| | return { |
| | "text": tokenizer.apply_chat_template( |
| | example["messages"], |
| | tokenize=False, |
| | ) |
| | } |
| | |
| | ds = ds.map(preprocess) |
| | |
| | # Tokenize inputs. |
| | def tokenize(sample): |
| | return tokenizer( |
| | sample["text"], |
| | padding=False, |
| | max_length=MAX_SEQUENCE_LENGTH, |
| | truncation=True, |
| | add_special_tokens=False, |
| | ) |
| | |
| | ds = ds.map(tokenize, remove_columns=ds.column_names) |
| | |
| | # Configure the quantization algorithm and scheme with explicit parameters. |
| | recipe = GPTQModifier( |
| | targets="Linear", |
| | scheme="W8A8", |
| | ignore=[ |
| | "lm_head", |
| | "re:.*mlp.gate$" |
| | ], |
| | ) |
| | |
| | # Apply quantization. |
| | oneshot( |
| | model=model, |
| | dataset=ds, |
| | recipe=recipe, |
| | max_seq_length=MAX_SEQUENCE_LENGTH, |
| | num_calibration_samples=NUM_CALIBRATION_SAMPLES, |
| | pipeline="sequential", |
| | sequential_targets=["Glm4MoeDecoderLayer"], |
| | trust_remote_code_model=True, |
| | ) |
| | |
| | SAVE_DIR = "./" + MODEL_ID.rstrip("/").split("/")[-1] + "-quantized.w8a8" |
| | model.save_pretrained(SAVE_DIR, save_compressed=True) |
| | tokenizer.save_pretrained(SAVE_DIR) |
| | |
| | ``` |
| | </details> |
| |
|
| | ## Evaluation |
| |
|
| | This model was evaluated on the well-known text benchmarks using [lm-evaluation-harness](https://github.com/neuralmagic/lm-evaluation-harness). The Reasoning evals were done using [ligheval](https://github.com/neuralmagic/lighteval). |
| |
|
| | ### Accuracy |
| |
|
| | <table> |
| | <thead> |
| | <tr> |
| | <th>Category</th> |
| | <th>Metric</th> |
| | <th>zai-org/GLM-4.6-FP8</th> |
| | <th>RedHatAI/GLM-4.6-quantized.w8a8 (this model)</th> |
| | <th>Recovery</th> |
| | </tr> |
| | </thead> |
| | <tbody> |
| | <!-- OpenLLM V1 --> |
| | <tr> |
| | <td rowspan="2"><b>Leaderboard</b></td> |
| | <td>MMLU Pro</td> |
| | <td>50.65%</td> |
| | <td>50.08%</td> |
| | <td>98.87%</td> |
| | </tr> |
| | <tr> |
| | <td>IFEVAL</td> |
| | <td>91.97%</td> |
| | <td>93.68%</td> |
| | <td>101.86%</td> |
| | </tr> |
| | <tr> |
| | <td rowspan="6"><b>Reasoning</b></td> |
| | <td>AIME25</td> |
| | <td>96.67%</td> |
| | <td>90.00%</td> |
| | <td>93.10%</td> |
| | </tr> |
| | <tr> |
| | <td>Math-500 (0-shot)</td> |
| | <td>88.80%</td> |
| | <td>90.60%</td> |
| | <td>102.03%</td> |
| | </tr> |
| | <tr> |
| | <td>GPQA (Diamond, 0-shot)</td> |
| | <td>81.82%</td> |
| | <td>78.78%</td> |
| | <td>96.28%</td> |
| | </tr> |
| | </tbody> |
| | </table> |
| | |
| |
|
| | ### Reproduction |
| |
|
| | The results were obtained using the following commands: |
| |
|
| | <details> |
| |
|
| | #### Leaderboard |
| |
|
| | ``` |
| | lm_eval --model local-chat-completions \ |
| | --tasks mmlu_pro \ |
| | --model_args "model=RedHatAI/GLM-4.6-quantized.w8a8,max_length=90000,base_url=http://0.0.0.0:3758/v1/chat/completions,num_concurrent=128,max_retries=3,tokenized_requests=False,tokenizer_backend=None,timeout=1200" \ |
| | --num_fewshot 5 \ |
| | --apply_chat_template \ |
| | --fewshot_as_multiturn \ |
| | --output_path ./ \ |
| | --seed 42 \ |
| | --gen_kwargs "do_sample=True,temperature=1.0,top_p=0.95,max_gen_toks=64000" |
| | |
| | |
| | lm_eval --model local-chat-completions \ |
| | --tasks leaderboard_ifeval \ |
| | --model_args "model=RedHatAI/GLM-4.6-quantized.w8a8,max_length=90000,base_url=http://0.0.0.0:3758/v1/chat/completions,num_concurrent=128,max_retries=3,tokenized_requests=False,tokenizer_backend=None,timeout=1200" \ |
| | --num_fewshot 5 \ |
| | --apply_chat_template \ |
| | --fewshot_as_multiturn \ |
| | --output_path ./ \ |
| | --seed 42 \ |
| | --gen_kwargs "do_sample=True,temperature=1.0,top_p=0.95,max_gen_toks=64000" |
| | ``` |
| |
|
| |
|
| | #### Reasoning |
| | ``` |
| | litellm_config.yaml: |
| | |
| | model_parameters: |
| | provider: "hosted_vllm" |
| | model_name: "hosted_vllm/redhatai-glm-4.6-w8a8" |
| | base_url: "http://0.0.0.0:3759/v1" |
| | api_key: "" |
| | timeout: 3600 |
| | concurrent_requests: 128 |
| | generation_parameters: |
| | temperature: 1.0 |
| | max_new_tokens: 131072 |
| | top_p: 0.95 |
| | seed: 0 |
| | |
| | lighteval endpoint litellm litellm_config.yaml \ |
| | "aime25|0,math_500|0,gpqa:diamond|0" \ |
| | --output-dir ./ \ |
| | --save-details |
| | ``` |
| |
|
| | </details> |
