| { | |
| "seed": 42, | |
| "output_dir": "/scratch/hf_st_mm_outputs/server_datacenter_8gpu_tri_encoder", | |
| "model": { | |
| "text_encoder_name": "llm-semantic-router/mmbert-embed-32k-2d-matryoshka", | |
| "image_encoder_name": "google/siglip2-so400m-patch14-384", | |
| "audio_encoder_name": "openai/whisper-medium", | |
| "embedding_dim": 768, | |
| "max_text_length": 32768 | |
| }, | |
| "training": { | |
| "epochs": 10, | |
| "batch_size": 12, | |
| "grad_accum_steps": 8, | |
| "num_workers": 4, | |
| "prefetch_factor": 4, | |
| "shard_prefetch": 2, | |
| "shard_cache_limit": 4, | |
| "sequential_shard_loading": true, | |
| "shuffle": false, | |
| "modality_homogeneous_batches": false, | |
| "learning_rate": 1e-05, | |
| "weight_decay": 0.01, | |
| "warmup_ratio": 0.1, | |
| "max_grad_norm": 1.0, | |
| "mixed_precision": "bf16", | |
| "log_every": 10, | |
| "save_every": 2000, | |
| "hard_negative_ratio": 0.5 | |
| }, | |
| "loss": { | |
| "type": "cached_mnrl", | |
| "scale": 20.0 | |
| }, | |
| "data": { | |
| "cache_dir": "/scratch/2dmse-data/server_full_datacenter_cache/train" | |
| }, | |
| "validation": { | |
| "cache_dir": "/scratch/2dmse-data/server_full_datacenter_cache/val", | |
| "num_workers": 2, | |
| "shard_prefetch": 1, | |
| "shard_cache_limit": 2 | |
| } | |
| } |