Feature Extraction
Transformers
Safetensors
English
tridlm
masked-diffusion
diffusion-language-model
gpt2
openwebtext
triangular-attention
custom_code
Instructions to use EER6/TriDLM-124M-split with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use EER6/TriDLM-124M-split with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("feature-extraction", model="EER6/TriDLM-124M-split", trust_remote_code=True)# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("EER6/TriDLM-124M-split", trust_remote_code=True, device_map="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "arm": "split", | |
| "attn_impl": "flex", | |
| "qk_norm": true, | |
| "output_dir": "/scratch/11079/antonxue/tridlm-runs/20260829-1047-qk1-lr1e-4-split", | |
| "learning_rate": 0.0001, | |
| "warmup_steps": 1000, | |
| "stable_steps": 89000, | |
| "decay_steps": 10000, | |
| "min_lr_ratio": 0.0, | |
| "weight_decay": 0.1, | |
| "batch_size": 64, | |
| "gradient_accumulation_steps": 1, | |
| "logging_steps": 5, | |
| "save_steps": 1000, | |
| "save_total_limit": 5, | |
| "keep_every": 5000, | |
| "seed": 1, | |
| "dataloader_num_workers": 4, | |
| "compile": true, | |
| "resume_from_checkpoint": null, | |
| "tiny": false, | |
| "synthetic_rows": 0, | |
| "max_train_rows": 0 | |
| } |