Safetensors
idefics3
ocr
code
math
formula
CodeFormulaV2 / config.json
MatteoOmenetti's picture
Update config.json
ecedbe1 verified
{
"_flash_attn_2_enabled": true,
"architectures": [
"Idefics3ForConditionalGeneration"
],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100264,
"eos_token_id": 100338,
"freeze_lm_head": true,
"freeze_text_layers": true,
"freeze_text_module_exceptions": [],
"freeze_vision_layers": true,
"freeze_vision_module_exceptions": [],
"head_dim": 64,
"hidden_act": "silu",
"hidden_size": 576,
"image_token_id": 100270,
"initializer_range": 0.02,
"intermediate_size": 1536,
"max_position_embeddings": 8192,
"mlp_bias": false,
"model_type": "idefics3",
"neftune_noise_alpha": 0.0,
"num_attention_heads": 9,
"num_hidden_layers": 30,
"num_key_value_heads": 3,
"pad_token_id": 100256,
"perceiver_config": {
"attention_dropout": 0.0,
"hidden_act": "silu",
"model_type": "vllama3",
"num_key_value_heads": 1,
"qk_layer_norms_perceiver": false,
"resampler_depth": 6,
"resampler_head_dim": 96,
"resampler_n_heads": 16,
"resampler_n_latents": 64
},
"pixel_shuffle_factor": 4,
"pretraining_tp": 1,
"qk_layer_norms": false,
"rms_norm_eps": 1e-05,
"rope_scaling": null,
"rope_theta": 100000.0,
"scale_factor": 4,
"text_config": {
"architectures": ["VLlama3ForCausalLM"],
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 100257,
"eos_token_id": 100257,
"head_dim": 64,
"hidden_act": "silu",
"hidden_size": 576,
"initializer_range": 0.02,
"intermediate_size": 1536,
"max_position_embeddings": 8192,
"mlp_bias": false,
"model_type": "llama",
"num_attention_heads": 9,
"num_hidden_layers": 30,
"num_key_value_heads": 3,
"pretraining_tp": 1,
"rms_norm_eps": 1e-5,
"rope_scaling": null,
"rope_theta": 10000.0,
"torch_dtype": "bfloat16",
"use_cache": true,
"vocab_size": 100480,
"tie_word_embeddings": false
},
"tie_word_embeddings": true,
"torch_dtype": "bfloat16",
"transformers_version": "4.51.3",
"use_cache": false,
"use_resampler": false,
"vision_config": {
"attention_dropout": 0.0,
"hidden_act": "gelu_pytorch_tanh",
"hidden_size": 768,
"image_size": 512,
"initializer_range": 0.02,
"intermediate_size": 3072,
"layer_norm_eps": 1e-06,
"max_image_size": {
"longest_edge": 512
},
"model_type": "idefics3_vision",
"num_attention_heads": 12,
"num_channels": 3,
"num_hidden_layers": 12,
"patch_size": 16,
"size": {
"longest_edge": 2048
},
"torch_dtype": "bfloat16",
"use_base_siglip": true
},
"vocab_size": 100480
}