fwerkor/MiniCPM5-2B-Diffusion-Base
2751
1{2 "format": "cid-diffusion-base-v1",3 "base_model": "openbmb/MiniCPM5-2B-Base",4 "objective": "LLaDA-style masked diffusion",5 "attention": "bidirectional",6 "mask_token": "<|cid_mask|>",7 "mask_token_id": 130560,8 "sequence_length": 2048,9 "completed_steps": 5087,10 "global_batch_size": 96,11 "tokens_seen": 1000144896,12 "training_corpus": {13 "format": "cid-diffusion-stream-v1",14 "base_model": "openbmb/MiniCPM5-2B-Base",15 "sequence_length": 2048,16 "sources": [17 {18 "name": "ultrax-web-en",19 "repo": "openbmb/UltraX-Preview",20 "revision": "a88527587389fd4ab352e9ad1273f4c0a234d8df",21 "pattern": "data/UltraX-Ultra-FineWeb/*.parquet",22 "column": "cleaned_content",23 "weight": 0.724 },25 {26 "name": "ultrafineweb-zh",27 "repo": "openbmb/Ultra-FineWeb",28 "revision": "02c85641e3d19a854be2e09139c25adaa9518063",29 "pattern": "data/ultrafineweb_zh/*.parquet",30 "column": "content",31 "weight": 0.1532 },33 {34 "name": "ultradata-code",35 "repo": "openbmb/UltraData-Code",36 "revision": "85182d829f2ce7ea07cca72ebfc509deea1d9f5f",37 "pattern": "data/UltraData-Code-L2/*/*.parquet",38 "column": "content",39 "weight": 0.140 },41 {42 "name": "ultradata-math",43 "repo": "openbmb/UltraData-Math",44 "revision": "fe10db8efd35597fd7fcff8ff576b5ec4ea5ff87",45 "pattern": "data/UltraData-Math-L2-preview/*.parquet",46 "column": "content",47 "weight": 0.0548 }49 ]50 },51 "mask_ratio_range": [52 0.001,53 1.054 ],55 "hf_loader": "CIDDiffusionForMaskedLM",56 "hf_auto_model": "AutoModelForCausalLM",57 "requires_trust_remote_code": true58}59 