tinygemma3ocr2m / hf /config.json
shibatch's picture
Upload folder using huggingface_hub
86d0ebe verified
Raw
History Blame Contribute Delete
1.94 kB
{
"architectures": [
"Gemma3ForConditionalGeneration"
],
"boi_token_index": 255999,
"dtype": "bfloat16",
"eoi_token_index": 256000,
"image_token_index": 1003,
"initializer_range": 0.02,
"mm_tokens_per_image": 64,
"model_type": "gemma3",
"text_config": {
"_sliding_window_pattern": 6,
"attention_bias": false,
"attention_dropout": 0.0,
"attn_logit_softcapping": null,
"bos_token_id": 1000,
"dtype": "bfloat16",
"eos_token_id": 1001,
"final_logit_softcapping": null,
"head_dim": 32,
"hidden_activation": "gelu_pytorch_tanh",
"hidden_size": 128,
"initializer_range": 0.02,
"intermediate_size": 512,
"layer_types": [
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"sliding_attention",
"full_attention"
],
"max_position_embeddings": 256,
"model_type": "gemma3_text",
"num_attention_heads": 4,
"num_hidden_layers": 6,
"num_key_value_heads": 1,
"pad_token_id": 1000,
"query_pre_attn_scalar": 32,
"rms_norm_eps": 1e-06,
"rope_parameters": {
"full_attention": {
"rope_theta": 1000000.0,
"rope_type": "default"
},
"sliding_attention": {
"rope_theta": 10000.0,
"rope_type": "default"
}
},
"sliding_window": 32,
"tie_word_embeddings": true,
"use_bidirectional_attention": false,
"use_cache": false,
"vocab_size": 1024
},
"tie_word_embeddings": true,
"transformers_version": "5.9.0",
"vision_config": {
"attention_dropout": 0.0,
"dtype": "bfloat16",
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 128,
"image_size": 128,
"intermediate_size": 256,
"layer_norm_eps": 1e-06,
"model_type": "siglip_vision_model",
"num_attention_heads": 4,
"num_channels": 3,
"num_hidden_layers": 2,
"patch_size": 16
},
"vocab_size": 1024
}