Instructions to use tencent/HY-MT1.5-7B-GPTQ-Int4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use tencent/HY-MT1.5-7B-GPTQ-Int4 with Transformers:
# Use a pipeline as a high-level helper # Warning: Pipeline type "translation" is no longer supported in transformers v5. # You must load the model directly (see below) or downgrade to v4.x with: # 'pip install "transformers<5.0.0' from transformers import pipeline pipe = pipeline("translation", model="tencent/HY-MT1.5-7B-GPTQ-Int4")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("tencent/HY-MT1.5-7B-GPTQ-Int4") model = AutoModelForCausalLM.from_pretrained("tencent/HY-MT1.5-7B-GPTQ-Int4", device_map="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "model_config": { | |
| "name": "HunyuanDense", | |
| "model_path": "Base Model Path", | |
| "trust_remote_code": true, | |
| "torch_dtype": "auto", | |
| "device_map": "cpu", | |
| "low_cpu_mem_usage": true, | |
| "use_cache": false, | |
| "cache_dir": null, | |
| "use_audio_in_video": false, | |
| "attn_implementation": "default" | |
| }, | |
| "compression_config": { | |
| "name": [ | |
| "PTQ" | |
| ], | |
| "quantization": { | |
| "name": "int4_gptq", | |
| "save_name": "compressed-tensors", | |
| "bits": 4, | |
| "quant_method": { | |
| "weight": "per-group", | |
| "group_size": 64 | |
| }, | |
| "quant_helpers": [], | |
| "smooth_alpha": 0.5, | |
| "low_memory": false, | |
| "cpu_convert": false, | |
| "modules_to_quantize": [], | |
| "zero_point": true, | |
| "mse_range": false, | |
| "ignore_layers": [ | |
| "lm_head", | |
| "mlp.gate.wg" | |
| ], | |
| "quant_analyse": false, | |
| "quant_vit": false | |
| }, | |
| "cache": null | |
| }, | |
| "dataset_config": { | |
| "name": "TextDataset", | |
| "data_path": "Data Path", | |
| "max_seq_length": 4096, | |
| "num_samples": 1024, | |
| "batch_size": 1, | |
| "shuffle": false, | |
| "inference_settings": null | |
| }, | |
| "global_config": { | |
| "save_path": "Save Model Path", | |
| "max_seq_length": 4096, | |
| "hidden_size": 4096, | |
| "model_arch_type": "hunyuan_v1_dense", | |
| "absolute_model_path": "", | |
| "deploy_backend": "vllm" | |
| }, | |
| "infer_config": null, | |
| "debug_info": { | |
| "python": "3.10.14 (main, Mar 6 2025, 18:16:47) [GCC 11.4.0]", | |
| "angelslim": { | |
| "name": "angelslim", | |
| "version": "N/A", | |
| "source": "Unknown" | |
| }, | |
| "torch": { | |
| "name": "torch", | |
| "version": "2.7.0", | |
| "source": "pip" | |
| }, | |
| "transformers": { | |
| "name": "transformers", | |
| "version": "4.57.3", | |
| "source": "pip" | |
| }, | |
| "torch_cuda_version": "12.6" | |
| } | |
| } | |