trl-internal-testing/tiny-DiffusionGemmaForBlockDiffusion
325k
1{2 "architectures": [3 "DiffusionGemmaForBlockDiffusion"4 ],5 "boi_token_id": 255999,6 "canvas_length": 32,7 "dtype": "bfloat16",8 "eoi_token_id": 258882,9 "eos_token_id": [10 1,11 10612 ],13 "image_token_id": 258880,14 "initializer_range": 0.02,15 "model_type": "diffusion_gemma",16 "text_config": {17 "attention_bias": false,18 "attention_dropout": 0.0,19 "bos_token_id": 2,20 "dtype": "bfloat16",21 "eos_token_id": 1,22 "final_logit_softcapping": 30.0,23 "global_head_dim": 8,24 "head_dim": 8,25 "hidden_activation": "gelu_pytorch_tanh",26 "hidden_size": 16,27 "initializer_range": 0.02,28 "intermediate_size": 32,29 "layer_types": [30 "sliding_attention",31 "full_attention"32 ],33 "max_position_embeddings": 262144,34 "model_type": "diffusion_gemma_text",35 "moe_intermediate_size": 16,36 "num_attention_heads": 4,37 "num_experts": 4,38 "num_global_key_value_heads": 2,39 "num_hidden_layers": 2,40 "num_key_value_heads": 2,41 "pad_token_id": 0,42 "rms_norm_eps": 1e-06,43 "rope_parameters": {44 "full_attention": {45 "partial_rotary_factor": 0.25,46 "rope_theta": 1000000.0,47 "rope_type": "proportional"48 },49 "sliding_attention": {50 "rope_theta": 10000.0,51 "rope_type": "default"52 }53 },54 "sliding_window": 1024,55 "tie_word_embeddings": true,56 "top_k_experts": 2,57 "use_bidirectional_attention": "vision",58 "vocab_size": 26214459 },60 "tie_word_embeddings": true,61 "transformers_version": "5.10.0.dev0",62 "vision_config": {63 "_name_or_path": "",64 "architectures": null,65 "attention_bias": false,66 "attention_dropout": 0.0,67 "chunk_size_feed_forward": 0,68 "default_output_length": 70,69 "dtype": "bfloat16",70 "embed_dim": 64,71 "global_head_dim": 72,72 "head_dim": 72,73 "hidden_activation": "gelu_pytorch_tanh",74 "hidden_size": 16,75 "id2label": {76 "0": "LABEL_0",77 "1": "LABEL_1"78 },79 "initializer_range": 0.02,80 "intermediate_size": 32,81 "is_encoder_decoder": false,82 "label2id": {83 "LABEL_0": 0,84 "LABEL_1": 185 },86 "max_position_embeddings": 131072,87 "model_type": "gemma4_vision",88 "num_attention_heads": 4,89 "num_hidden_layers": 2,90 "num_key_value_heads": 2,91 "output_attentions": false,92 "output_hidden_states": false,93 "patch_size": 16,94 "pooling_kernel_size": 3,95 "position_embedding_size": 630,96 "problem_type": null,97 "return_dict": true,98 "rms_norm_eps": 1e-06,99 "rope_parameters": {100 "rope_theta": 100.0,101 "rope_type": "default"102 },103 "standardize": true,104 "use_clipped_linears": false105 },106 "vision_soft_tokens_per_image": 280107}108 