{ "model_type": "tiny_diffusion_gemma_moe_5m", "hf_model_type": "diffusion_gemma", "parameter_count": 4190512, "trainable_parameter_count": 4147360, "tokenizer_size": 1003, "split_sizes": { "ar_train_rows": 2098522, "ar_validation_rows": 21197, "diffusion_train_rows": 2098522, "diffusion_validation_rows": 21197 }, "packed_sizes": { "ar_train_blocks": 9764172, "ar_validation_blocks": 98619, "diffusion_train_blocks": 9764172, "diffusion_validation_blocks": 98619 }, "phase_metrics": { "phase_1_ar": {}, "phase_2_partial_diffusion": {}, "phase_3_full_diffusion": { "noise_2_of_8_stochastic_first_accuracy": 0.42646167799830437, "noise_2_of_8_stochastic_first_accuracy_full": 0.85546875, "noise_2_of_8_stochastic_first_loss": 2.6051205694675446, "noise_2_of_8_stochastic_first_loss_full": 0.7259416617453098, "noise_2_of_8_stochastic_second_accuracy": 0.4603654481470585, "noise_2_of_8_stochastic_second_accuracy_full": 0.857421875, "noise_2_of_8_stochastic_second_loss": 2.4278698414564133, "noise_2_of_8_stochastic_second_loss_full": 0.68489970266819, "noise_4_of_8_stochastic_first_accuracy": 0.3474049922078848, "noise_4_of_8_stochastic_first_accuracy_full": 0.66015625, "noise_4_of_8_stochastic_first_loss": 3.226193368434906, "noise_4_of_8_stochastic_first_loss_full": 1.7175433486700058, "noise_4_of_8_stochastic_second_accuracy": 0.3570522516965866, "noise_4_of_8_stochastic_second_accuracy_full": 0.66796875, "noise_4_of_8_stochastic_second_loss": 3.0200355648994446, "noise_4_of_8_stochastic_second_loss_full": 1.6022306233644485, "noise_6_of_8_stochastic_first_accuracy": 0.24033872783184052, "noise_6_of_8_stochastic_first_accuracy_full": 0.4140625, "noise_6_of_8_stochastic_first_loss": 4.164743721485138, "noise_6_of_8_stochastic_first_loss_full": 3.2314010858535767, "noise_6_of_8_stochastic_second_accuracy": 0.2592960111796856, "noise_6_of_8_stochastic_second_accuracy_full": 0.4296875, "noise_6_of_8_stochastic_second_loss": 4.050312131643295, "noise_6_of_8_stochastic_second_loss_full": 3.140813171863556, "noise_8_of_8_stochastic_first_accuracy": 0.130859375, "noise_8_of_8_stochastic_first_accuracy_full": 0.130859375, "noise_8_of_8_stochastic_first_loss": 4.9496771693229675, "noise_8_of_8_stochastic_first_loss_full": 4.9496771693229675, "noise_8_of_8_stochastic_second_accuracy": 0.142578125, "noise_8_of_8_stochastic_second_accuracy_full": 0.142578125, "noise_8_of_8_stochastic_second_loss": 4.875990211963654, "noise_8_of_8_stochastic_second_loss_full": 4.875990211963654 } }, "final_ar_validation_loss": 2.0182483792304993, "final_ar_validation_ppl": 7.525132204842911, "final_diffusion_metrics": { "noise_2_of_8_stochastic_first_accuracy": 0.42646167799830437, "noise_2_of_8_stochastic_first_accuracy_full": 0.85546875, "noise_2_of_8_stochastic_first_loss": 2.6051205694675446, "noise_2_of_8_stochastic_first_loss_full": 0.7259416617453098, "noise_2_of_8_stochastic_second_accuracy": 0.4603654481470585, "noise_2_of_8_stochastic_second_accuracy_full": 0.857421875, "noise_2_of_8_stochastic_second_loss": 2.4278698414564133, "noise_2_of_8_stochastic_second_loss_full": 0.68489970266819, "noise_4_of_8_stochastic_first_accuracy": 0.3474049922078848, "noise_4_of_8_stochastic_first_accuracy_full": 0.66015625, "noise_4_of_8_stochastic_first_loss": 3.226193368434906, "noise_4_of_8_stochastic_first_loss_full": 1.7175433486700058, "noise_4_of_8_stochastic_second_accuracy": 0.3570522516965866, "noise_4_of_8_stochastic_second_accuracy_full": 0.66796875, "noise_4_of_8_stochastic_second_loss": 3.0200355648994446, "noise_4_of_8_stochastic_second_loss_full": 1.6022306233644485, "noise_6_of_8_stochastic_first_accuracy": 0.24033872783184052, "noise_6_of_8_stochastic_first_accuracy_full": 0.4140625, "noise_6_of_8_stochastic_first_loss": 4.164743721485138, "noise_6_of_8_stochastic_first_loss_full": 3.2314010858535767, "noise_6_of_8_stochastic_second_accuracy": 0.2592960111796856, "noise_6_of_8_stochastic_second_accuracy_full": 0.4296875, "noise_6_of_8_stochastic_second_loss": 4.050312131643295, "noise_6_of_8_stochastic_second_loss_full": 3.140813171863556, "noise_8_of_8_stochastic_first_accuracy": 0.130859375, "noise_8_of_8_stochastic_first_accuracy_full": 0.130859375, "noise_8_of_8_stochastic_first_loss": 4.9496771693229675, "noise_8_of_8_stochastic_first_loss_full": 4.9496771693229675, "noise_8_of_8_stochastic_second_accuracy": 0.142578125, "noise_8_of_8_stochastic_second_accuracy_full": 0.142578125, "noise_8_of_8_stochastic_second_loss": 4.875990211963654, "noise_8_of_8_stochastic_second_loss_full": 4.875990211963654 }, "final_ema_ar_validation_loss": 2.014301896095276, "final_ema_ar_validation_ppl": 7.4954929214408565, "final_ema_diffusion_metrics": { "noise_2_of_8_stochastic_first_accuracy": 0.42535873129963875, "noise_2_of_8_stochastic_first_accuracy_full": 0.85546875, "noise_2_of_8_stochastic_first_loss": 2.603432834148407, "noise_2_of_8_stochastic_first_loss_full": 0.7258147932589054, "noise_2_of_8_stochastic_second_accuracy": 0.47312964498996735, "noise_2_of_8_stochastic_second_accuracy_full": 0.861328125, "noise_2_of_8_stochastic_second_loss": 2.423415556550026, "noise_2_of_8_stochastic_second_loss_full": 0.6831153854727745, "noise_4_of_8_stochastic_first_accuracy": 0.3335161041468382, "noise_4_of_8_stochastic_first_accuracy_full": 0.658203125, "noise_4_of_8_stochastic_first_loss": 3.2337255477905273, "noise_4_of_8_stochastic_first_loss_full": 1.7171394303441048, "noise_4_of_8_stochastic_second_accuracy": 0.3573715351521969, "noise_4_of_8_stochastic_second_accuracy_full": 0.671875, "noise_4_of_8_stochastic_second_loss": 3.0173322558403015, "noise_4_of_8_stochastic_second_loss_full": 1.5972065776586533, "noise_6_of_8_stochastic_first_accuracy": 0.24319838173687458, "noise_6_of_8_stochastic_first_accuracy_full": 0.416015625, "noise_6_of_8_stochastic_first_loss": 4.165034174919128, "noise_6_of_8_stochastic_first_loss_full": 3.231293201446533, "noise_6_of_8_stochastic_second_accuracy": 0.2591778077185154, "noise_6_of_8_stochastic_second_accuracy_full": 0.431640625, "noise_6_of_8_stochastic_second_loss": 4.044229626655579, "noise_6_of_8_stochastic_second_loss_full": 3.13301882147789, "noise_8_of_8_stochastic_first_accuracy": 0.12890625, "noise_8_of_8_stochastic_first_accuracy_full": 0.12890625, "noise_8_of_8_stochastic_first_loss": 4.936682164669037, "noise_8_of_8_stochastic_first_loss_full": 4.936682164669037, "noise_8_of_8_stochastic_second_accuracy": 0.140625, "noise_8_of_8_stochastic_second_accuracy_full": 0.140625, "noise_8_of_8_stochastic_second_loss": 4.863627254962921, "noise_8_of_8_stochastic_second_loss_full": 4.863627254962921 }, "normal_generations": { "Once upon": "Once upon a time, there was a happy voice. He was very kindful and calm. He had made his friends who wrap them the time. One day, he saw a new geater in his bright flower. He thought a special game, so it could dry", "There was a little": "There was a little girl named Lily. She liked to play with her toys and spend all her things inls with her friends. One day, Lily's.", "One day": "One day, a big harage named Tim wanted to play near it. He saw a big mottle in the lard. He wanted to say sorry to his friend, \"Rrue, Tim. He ruffed you. He was very curious. He looked around his play" }, "ema_generations": { "Once upon": "Once upon a time, there was a little girl. She loved to play on the swings and play. One day, she decided to go.", "There was a little": "There was a little girl named Lily. She liked to play with her toys and spend all her toys in her room with her mommy. One day, she went to play with her toys and Lily saw a beautiful yellow berriy. She wanted to make it, but her mom said her it was right. Her", "One day": "One day, a little girl named Mia and Ben went walking in the park. They saw a big block and a big car. Lily wanted to toub it, but she just wanted it.\n\n\"Oh, Ben, what is that, doing?\" Ben asked.\n\n\"Well," }, "args": { "output_dir": "runs/diffusion_gemma_100p_full220k_resume80k", "init_model_dir": "runs/diffusion_gemma_100p_scratch_full220k/checkpoints/full-0080000/hf", "train_files": [ "data/train-00000-of-00004-2d5a1467fff1081b.parquet", "data/train-00001-of-00004-5852b56a2bd28fd9.parquet", "data/train-00002-of-00004-a26307300439e943.parquet", "data/train-00003-of-00004-d243063613e5a057.parquet" ], "max_rows": 0, "data_fraction": 1.0, "validation_fraction": 0.01, "disjoint_phase_data": false, "vocab_size": 1024, "base_vocab_size": 1000, "min_frequency": 2, "legacy_tokenizer": true, "legacy_special_token_ids": true, "hidden_size": 128, "intermediate_size": 384, "moe_intermediate_size": 192, "num_hidden_layers": 8, "num_attention_heads": 4, "num_key_value_heads": 1, "num_global_key_value_heads": 1, "head_dim": 32, "global_head_dim": 32, "num_experts": 4, "top_k_experts": 2, "sliding_window": 64, "max_position_embeddings": 1024, "layer_pattern": "sssssFsF", "canvas_length": 8, "max_parameters": 5000000, "vision_hidden_size": 32, "vision_intermediate_size": 64, "vision_num_hidden_layers": 1, "vision_num_attention_heads": 4, "vision_head_dim": 8, "vision_position_embedding_size": 64, "block_size": 64, "batch_size": 8, "gradient_accumulation_steps": 2, "ar_steps": 0, "partial_diffusion_steps": 0, "full_diffusion_steps": 140000, "ar_learning_rate": 0.0002, "diffusion_learning_rate": 7.41e-05, "minimum_learning_rate": 1e-05, "ar_warmup_steps": 1000, "partial_warmup_steps": 500, "full_warmup_steps": 0, "adam_beta1": 0.9, "adam_beta2": 0.95, "adam_eps": 1e-08, "weight_decay": 0.01, "grad_clip": 1.0, "error_on_nonfinite_gradients": true, "first_pass_loss_weight": 1.0, "clean_position_loss_weight": 1.0, "partial_ar_aux_weight": 1.0, "full_ar_aux_weight": 1.0, "partial_self_conditioning_probability": 1.0, "full_self_conditioning_probability": 1.0, "separate_noise_for_second_pass": false, "ema_decay": 0.999, "min_corruption_rate": 0.0, "max_corruption_rate": 1.0, "eval_stochastic_corruption": true, "log_steps": 50, "eval_steps": 1000, "eval_batches": 8, "save_steps": 10000, "seed": 1234, "device": "cuda", "dtype": "float32", "num_workers": 0, "generate_prompts": [ "Once upon", "There was a little", "One day" ], "generate_new_tokens": 64, "max_denoising_steps": 64, "entropy_bound": 1.0, "t_min": 0.4, "t_max": 0.8, "stability_threshold": 3, "confidence_threshold": 0.05 }, "config": { "transformers_version": "5.14.1", "architectures": [ "DiffusionGemmaForBlockDiffusion" ], "output_hidden_states": false, "return_dict": true, "dtype": "float32", "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": "float32", "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": 1003, "hidden_size": 128, "intermediate_size": 384, "num_hidden_layers": 8, "num_attention_heads": 4, "num_key_value_heads": 1, "head_dim": 32, "hidden_activation": "gelu_pytorch_tanh", "max_position_embeddings": 1024, "initializer_range": 0.02, "rms_norm_eps": 1e-06, "pad_token_id": 1000, "eos_token_id": 1001, "bos_token_id": 1000, "tie_word_embeddings": true, "rope_parameters": { "full_attention": { "partial_rotary_factor": 0.25, "rope_theta": 1000000.0, "rope_type": "proportional" }, "sliding_attention": { "rope_theta": 10000.0, "rope_type": "default" } }, "attention_bias": false, "attention_dropout": 0.0, "sliding_window": 64, "layer_types": [ "sliding_attention", "sliding_attention", "sliding_attention", "sliding_attention", "sliding_attention", "full_attention", "sliding_attention", "full_attention" ], "use_bidirectional_attention": "vision", "num_global_key_value_heads": 1, "global_head_dim": 32, "num_experts": 4, "top_k_experts": 2, "moe_intermediate_size": 192, "_name_or_path": "", "model_type": "diffusion_gemma_text", "output_attentions": false }, "vision_config": { "architectures": null, "output_hidden_states": false, "return_dict": true, "dtype": "float32", "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, "hidden_size": 32, "intermediate_size": 64, "num_hidden_layers": 1, "num_attention_heads": 4, "num_key_value_heads": 4, "head_dim": 8, "hidden_activation": "gelu_pytorch_tanh", "rms_norm_eps": 1e-06, "max_position_embeddings": 256, "attention_bias": false, "attention_dropout": 0.0, "rope_parameters": { "rope_theta": 100.0, "rope_type": "default" }, "pooling_kernel_size": 1, "patch_size": 16, "position_embedding_size": 64, "use_clipped_linears": false, "standardize": false, "initializer_range": 0.02, "_name_or_path": "", "model_type": "gemma4_vision", "output_attentions": false }, "boi_token_id": 255999, "eoi_token_id": 258882, "image_token_id": 258880, "initializer_range": 0.02, "tie_word_embeddings": true, "canvas_length": 8, "_name_or_path": "/home/codex/tinyModels/runs/diffusion_gemma_100p_scratch_full220k/checkpoints/full-0080000/hf", "model_type": "diffusion_gemma", "output_attentions": false }, "generation_config": { "max_new_tokens": 64, "max_length": null, "return_dict_in_generate": true, "max_denoising_steps": 64, "sampler_config": { "entropy_bound": 1.0, "_cls_name": "EntropyBoundSamplerConfig" }, "t_min": 0.4, "t_max": 0.8, "stability_threshold": 3, "confidence_threshold": 0.05, "cache_implementation": "dynamic", "cache_config": null, "disable_compile": null, "bos_token_id": 1000, "pad_token_id": 1000, "eos_token_id": 1001, "_from_model_config": null, "transformers_version": "5.14.1" } }