MINT-light-zero / config.json
huangrm's picture
Upload MINT-light-zero checkpoint
fae37da verified
Raw History Blame Contribute Delete
2.82 kB
{
"type": "mint_light",
"n_obs_steps": 1,
"input_features": {
"observation.state": {
"type": "STATE",
"shape": [
8
]
},
"observation.images.image": {
"type": "VISUAL",
"shape": [
3,
128,
128
]
},
"observation.images.image2": {
"type": "VISUAL",
"shape": [
3,
128,
128
]
}
},
"output_features": {
"action": {
"type": "ACTION",
"shape": [
7
]
}
},
"device": "cuda",
"use_amp": false,
"use_peft": false,
"push_to_hub": false,
"repo_id": null,
"private": null,
"tags": null,
"license": null,
"pretrained_path": null,
"chunk_size": 16,
"action_history_steps": 0,
"n_action_steps": 4,
"max_state_dim": 32,
"max_action_dim": 7,
"image_resolution": [
224,
224
],
"hidden_dim": 384,
"num_layers": 10,
"num_heads": 6,
"head_dim": 64,
"mlp_hidden_dim": 1024,
"attention_dropout": 0.0,
"ffn_dropout": 0.05,
"drop_path_rate": 0.0,
"cross_attention_every": 2,
"vision_backbone_id": "dinov3-vit-l-224px",
"vision_pretrained": true,
"use_language": false,
"language_model_name": "google/siglip2-so400m-patch14-224",
"language_pretrained": true,
"tokenizer_max_length": 64,
"state_hidden_dim": 128,
"start_query_hidden_dim": 384,
"classifier_hidden_dim": 384,
"ema_decay": 0.999,
"sample_top_k": 3,
"intent_ensemble": true,
"intent_ensemble_temperature": 0.5,
"intent_ensemble_decay": 0.03,
"codebook_size": 512,
"codebook_dim": 32,
"patch_nums": [
1,
2,
4
],
"vqvae_name_or_path": "/inspire/qb-ilm/project/robot-decision/huangrenming-253108120148/MINT/outputs/sdat/h_16_bs_256_cbs_512_cbd_32_ch_48/2026-06-19-03-12-57/vae-ckpt-40.pth",
"vqvae_ch": 48,
"vqvae_ch_mult": [
2,
4,
8
],
"vqvae_dropout": 0.1,
"vqvae_beta": 0.25,
"vqvae_using_znorm": true,
"vqvae_quant_conv_ks": 3,
"vqvae_quant_resi": 0.5,
"vqvae_share_quant_resi": 0,
"vqvae_patchwise": {
"enable": true,
"d_embed": 8,
"grouped_depth": 2,
"norm": "layer"
},
"optimizer_lr": 0.0003,
"optimizer_weight_decay": 0.005,
"optimizer_grad_clip_norm": 2.0,
"scheduler_warmup_steps": 1000,
"scheduler_decay_steps": 100000,
"scheduler_decay_lr": 9e-05,
"normalization_mapping": {
"VISUAL": "IDENTITY",
"STATE": "QUANTILES",
"ACTION": "IDENTITY"
}
}