Instructions to use YanZhanPKU/dLLM-PRM-Gap-bidir-dream7b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- PEFT
How to use YanZhanPKU/dLLM-PRM-Gap-bidir-dream7b with PEFT:
Task type is invalid.
- Notebooks
- Google Colab
- Kaggle
File size: 990 Bytes
3c14026 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 | {
"base_model": "Dream-org/Dream-v0-Instruct-7B",
"adapter_type": "LoRA plus reward head",
"attention": "bidirectional",
"pool_strategy": "mean",
"mask_aware_pooling": true,
"step_embedding": true,
"lora": {
"r": 16,
"alpha": 32,
"dropout": 0.05,
"target_modules": [
"q_proj",
"v_proj"
]
},
"release_status": "public release",
"checkpoint_identity_status": "exact submitted-paper main PRM",
"paper_role": "process-reward model used by PRM Guided, Hybrid, and the snapshot diagnostics",
"adapter_bytes": 36202468,
"parameter_prefixes_kept": [
"lora_A",
"lora_B",
"reward_head",
"step_proj",
"step_embed"
],
"causal": false,
"no_step_embed": false,
"no_mask_aware": false,
"lora_r": 16,
"lora_alpha": 32,
"lora_dropout": 0.05,
"step_embed_dim": 256,
"reward_hidden": 1024,
"seed": 42,
"max_steps": 3000,
"best_step": 2500,
"train_samples": 1276560,
"validation_samples_sampled": 6000
}
|