Robometer-4B / config.json
lilkm's picture
lilkm HF Staff
Upload folder using huggingface_hub
db167a7 verified
{
"type": "robometer",
"input_features": {
"observation.images.top": {
"type": "VISUAL",
"shape": [
3,
224,
224
]
}
},
"output_features": {
"progress": {
"type": "REWARD",
"shape": [
1
]
},
"success": {
"type": "REWARD",
"shape": [
1
]
}
},
"device": "cpu",
"pretrained_path": "robometer/Robometer-4B",
"push_to_hub": false,
"repo_id": null,
"license": "apache-2.0",
"tags": [
"reward-model",
"vision-language",
"qwen3-vl",
"zero-shot"
],
"private": null,
"image_key": "observation.images.top",
"task_key": "task",
"default_task": null,
"max_frames": 8,
"reward_output": "progress",
"success_threshold": 0.5,
"base_model_id": "Qwen/Qwen3-VL-4B-Instruct",
"torch_dtype": "bfloat16",
"use_multi_image": true,
"use_per_frame_progress_token": true,
"average_temporal_patches": true,
"frame_pooling": "mean",
"frame_pooling_attn_temperature": 1.0,
"progress_loss_type": "discrete",
"progress_discrete_bins": 10,
"vlm_config": {
"transformers_version": "5.5.4",
"architectures": [
"Qwen3VLForConditionalGeneration"
],
"output_hidden_states": false,
"return_dict": true,
"dtype": null,
"chunk_size_feed_forward": 0,
"is_encoder_decoder": false,
"id2label": {
"0": "LABEL_0",
"1": "LABEL_1"
},
"label2id": {
"LABEL_0": 0,
"LABEL_1": 1
},
"problem_type": null,
"text_config": {
"architectures": null,
"output_hidden_states": false,
"return_dict": true,
"dtype": "bfloat16",
"chunk_size_feed_forward": 0,
"is_encoder_decoder": false,
"id2label": {
"0": "LABEL_0",
"1": "LABEL_1"
},
"label2id": {
"LABEL_0": 0,
"LABEL_1": 1
},
"problem_type": null,
"vocab_size": 151674,
"hidden_size": 2560,
"intermediate_size": 9728,
"num_hidden_layers": 36,
"num_attention_heads": 32,
"num_key_value_heads": 8,
"head_dim": 128,
"hidden_act": "silu",
"max_position_embeddings": 262144,
"initializer_range": 0.02,
"rms_norm_eps": 1e-06,
"use_cache": true,
"rope_parameters": {
"mrope_interleaved": true,
"mrope_section": [
24,
20,
20
],
"rope_type": "default",
"rope_theta": 5000000
},
"attention_bias": false,
"attention_dropout": 0.0,
"pad_token_id": null,
"_name_or_path": "",
"bos_token_id": 151643,
"eos_token_id": 151645,
"model_type": "qwen3_vl_text",
"tie_word_embeddings": true,
"output_attentions": false
},
"vision_config": {
"architectures": null,
"output_hidden_states": false,
"return_dict": true,
"dtype": null,
"chunk_size_feed_forward": 0,
"is_encoder_decoder": false,
"id2label": {
"0": "LABEL_0",
"1": "LABEL_1"
},
"label2id": {
"LABEL_0": 0,
"LABEL_1": 1
},
"problem_type": null,
"depth": 24,
"hidden_size": 1024,
"hidden_act": "gelu_pytorch_tanh",
"intermediate_size": 4096,
"num_heads": 16,
"in_channels": 3,
"patch_size": 16,
"spatial_merge_size": 2,
"temporal_patch_size": 2,
"out_hidden_size": 2560,
"num_position_embeddings": 2304,
"deepstack_visual_indexes": [
5,
11,
17
],
"initializer_range": 0.02,
"_name_or_path": "",
"model_type": "qwen3_vl",
"output_attentions": false
},
"image_token_id": 151655,
"video_token_id": 151656,
"vision_start_token_id": 151652,
"vision_end_token_id": 151653,
"tie_word_embeddings": true,
"_name_or_path": "Qwen/Qwen3-VL-4B-Instruct",
"model_type": "qwen3_vl",
"output_attentions": false
},
"normalization_mapping": {
"VISUAL": "IDENTITY",
"REWARD": "IDENTITY"
}
}