{ "fixture": "huggingface-mic/Inkling-mic", "key": "209ad3f38c33094dcca1347ed1ad26a7598cb9864074d672dec9bc1efd6b52a8", "key_fields": { "architectures": [ "InklingForConditionalGeneration" ], "output_hidden_states": false, "return_dict": true, "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": 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, "vocab_size": 201024, "unpadded_vocab_size": 200058, "hidden_size": 64, "num_hidden_layers": 3, "num_attention_heads": 4, "num_key_value_heads": 2, "head_dim": 16, "swa_num_attention_heads": 2, "swa_num_key_value_heads": 2, "swa_head_dim": 16, "sliding_window_size": 512, "d_rel": 16, "rel_extent": 1024, "log_scaling_n_floor": 128000, "log_scaling_alpha": 0.1, "local_layer_ids": [ 0, 1, 2, 3, 4, 6, 7, 8, 9, 10, 12, 13, 14, 15, 16, 18, 19, 20, 21, 22, 24, 25, 26, 27, 28, 30, 31, 32, 33, 34, 36, 37, 38, 39, 40, 42, 43, 44, 45, 46, 48, 49, 50, 51, 52, 54, 55, 56, 57, 58, 60, 61, 62, 63, 64 ], "layer_types": [ "hybrid_sliding", "hybrid_sliding", "hybrid" ], "max_position_embeddings": 1048576, "rms_norm_eps": 1e-06, "conv_kernel_size": 4, "mlp_layer_types": [ "dense", "sparse", "sparse" ], "intermediate_size": 128, "hidden_act": "silu", "moe_intermediate_size": 32, "n_routed_experts": 4, "num_experts_per_tok": 2, "n_shared_experts": 2, "shared_expert_sink": true, "route_scale": 8.0, "logits_mup_width_multiplier": 24.0, "rms_norm_eps_moe_gate": 1e-06, "attention_dropout": 0.0, "initializer_range": 0.02, "pad_token_id": null, "bos_token_id": 1, "eos_token_id": 2, "num_mtp_layers": 8, "chain_hidden_post_norm": false, "mtp_hidden_states_first": true, "mtp_local_layer_ids": [ 0, 2, 4, 5, 6, 7 ], "number_of_conv_states": 4, "_name_or_path": "", "q_bias": false, "o_bias": false, "use_embed_norm": true, "use_sconv": true, "final_logit_softcapping": null, "use_gate_bias": true, "gate_activation": "sigmoid", "norm_after_topk": true, "use_global_scale": true, "model_type": "inkling_text", "output_attentions": false }, "audio_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, "n_mel_bins": 80, "mel_vocab_size": 16, "text_hidden_size": 64, "rms_norm_eps": 1e-06, "initializer_range": 0.02, "_name_or_path": "", "decoder_dmodel": 6144, "bias": false, "dmel_min_value": -7.0, "dmel_max_value": 2.0, "use_audio_norm": true, "audio_mode": "dmel", "model_type": "inkling_audio", "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, "text_hidden_size": 64, "patch_size": 40, "temporal_patch_size": 2, "num_channels": 3, "hidden_size": 64, "n_layers": 2, "num_attention_heads": 4, "rms_norm_eps": 1e-06, "initializer_range": 0.02, "_name_or_path": "", "vision_encoder_type": "hmlp", "decoder_dmodel": 6144, "n_channels": 3, "use_vision_norm": true, "model_type": "inkling_vision", "output_attentions": false }, "image_token_id": 200054, "audio_token_id": 200053, "image_bos_token_id": 200005, "audio_bos_token_id": 200020, "model_type": "inkling_mm_model", "eos_token_id": 200006, "mtp_config": { "num_nextn_predict_layers": 8, "chain_hidden_post_norm": false, "local_layer_ids": [ 0, 2, 4, 5, 6, 7 ] }, "output_attentions": false }, "covers": { "thinkingmachines/Inkling": "828496eeae4c243ff1a22f7f28ff83694f2f7bc9" }, "built_from": { "repo": "thinkingmachines/Inkling", "revision": "828496eeae4c243ff1a22f7f28ff83694f2f7bc9" }, "seed": 17, "dtype": "float32", "builder": "example/build_fixtures.py", "builder_revision": "e8034a51dc828ccfb8c6ad3d765dd2275181f0b8", "recipe": { "seed": 17, "model_class": "InklingForConditionalGeneration" }, "shrink": "rule", "model_class": "InklingForConditionalGeneration", "parameters": 26736419, "missing_parameters": [], "serialization": { "file": "model.safetensors", "reverted_conversions": [ "WeightRenaming: ['model\\\\.llm\\\\.layers'] -> ['model.language_model.layers']", "WeightRenaming: ['model\\\\.llm\\\\.embed_norm\\\\.weight'] -> ['model.language_model.embed_tokens.embed_norm.weight']", "WeightRenaming: ['model\\\\.llm\\\\.embed\\\\.weight'] -> ['model.language_model.embed_tokens.weight']", "WeightRenaming: ['model\\\\.llm\\\\.norm\\\\.weight'] -> ['model.language_model.norm.weight']", "WeightRenaming: ['model\\\\.llm\\\\.unembed\\\\.weight'] -> ['lm_head.weight']", "WeightRenaming: ['model\\\\.audio\\\\.'] -> ['model.audio_tower.']", "WeightRenaming: ['model\\\\.visual'] -> ['model.vision_tower']", "WeightRenaming: ['vision_tower.layers.linear_(\\\\d+)'] -> ['vision_tower.encoder_layers.\\\\1.projection']", "WeightRenaming: ['vision_tower.layers.norm_(\\\\d+)'] -> ['vision_tower.encoder_layers.\\\\1.layer_norm']", "WeightRenaming: ['audio_tower.encoder.weight'] -> ['audio_tower.embed_audio_tokens.embed_audio_tokens.weight']", "WeightRenaming: ['audio_tower.final_norm.weight'] -> ['audio_tower.norm.weight']", "WeightConverter: ['shared_w13_weight'] -> ['gate_proj', 'up_proj']", "WeightRenaming: ['shared_w2_weight'] -> ['down_proj']", "WeightConverter: ['mlp.experts.w13_weight'] -> ['mlp.experts.gate_up_proj']", "WeightRenaming: ['mlp.experts.w2_weight'] -> ['mlp.experts.down_proj']", "WeightConverter: ['mlp.w13_dn.weight'] -> ['mlp.gate_proj.weight', 'mlp.up_proj.weight']", "WeightRenaming: ['mlp.w2_md.weight'] -> ['mlp.down_proj.weight']", "WeightRenaming: ['mlp.gate.bias'] -> ['mlp.gate.e_score_correction_bias']", "WeightRenaming: ['attn\\\\.wq_du'] -> ['self_attn.q_proj']", "WeightRenaming: ['attn\\\\.wk_dv'] -> ['self_attn.k_proj']", "WeightRenaming: ['attn\\\\.wv_dv'] -> ['self_attn.v_proj']", "WeightRenaming: ['attn\\\\.wr_du'] -> ['self_attn.r_proj']", "WeightRenaming: ['attn\\\\.wo_ud'] -> ['self_attn.o_proj']", "WeightRenaming: ['\\\\.attn\\\\.q_norm'] -> ['.self_attn.q_norm']", "WeightRenaming: ['\\\\.attn\\\\.k_norm'] -> ['.self_attn.k_norm']", "WeightRenaming: ['\\\\.attn\\\\.k_sconv'] -> ['.self_attn.k_sconv.conv1d']", "WeightRenaming: ['\\\\.attn\\\\.v_sconv'] -> ['.self_attn.v_sconv.conv1d']", "WeightRenaming: ['\\\\.attn\\\\.rel_logits_proj'] -> ['.self_attn.rel_logits_proj']", "WeightRenaming: ['attn_sconv\\\\.weight'] -> ['attn_sconv.conv1d.weight']", "WeightRenaming: ['mlp_sconv\\\\.weight'] -> ['mlp_sconv.conv1d.weight']", "WeightRenaming: ['mlp_norm'] -> ['post_attention_layernorm']", "WeightRenaming: ['attn_norm'] -> ['input_layernorm']" ] }, "tester_sizes": { "source": "https://raw.githubusercontent.com/huggingface/transformers/main/tests/models/inkling/test_modeling_inkling.py", "fields": { "text_config.swa_head_dim": 16, "text_config.swa_num_attention_heads": 2, "text_config.swa_num_key_value_heads": 2 } }, "checkpoint_inspection": "parameter headers/indexes; bounded binary fallback", "numeric_health_scope": "CPU float32/bfloat16/float16; autocast for low-precision losses", "captured": { "call": "training step", "inputs": [ "attention_mask", "input_ids", "labels" ], "class": "InklingForConditionalGeneration" }, "weights": "random", "gate": { "random": { "pass": true, "accepted_insensitive": {}, "threshold": 0.001, "deltas": { "input_ids:roll_batch": 0.05011682212352753, "input_ids:reverse": 0.04614920914173126, "attention_mask:all_ones": null }, "weakest": "input_ids:reverse", "no_perturbation_defined": [], "not_a_forward_parameter": [] } }, "numeric_health": { "float32": { "outputs_finite": true, "loss_finite": true, "gradients_finite": true, "finite": true }, "bfloat16": { "outputs_finite": true, "loss_finite": true, "gradients_finite": true, "finite": true }, "float16": { "outputs_finite": true, "loss_finite": true, "unscaled_gradients_finite": true, "finite": true } }, "copied_unchanged": [ "chat_template.jinja", "processor_config.json", "special_tokens_map.json", "tokenizer.json", "tokenizer_config.json" ], "environment": { "python": "3.12.12", "machine": "arm64", "torch": "2.14.1", "transformers": "5.22.0.dev0", "tokenizers": "0.23.3" }, "sha256": { "chat_template.jinja": "0aa1aa0c729d90176dcaa00c440c8faffca2957ffb2cc4b79456ee6d02bcf43b", "config.json": "967261d39b02fa42d82ca2dbd87b8bac1303db35a94ae43c8bb134694e91272e", "generation_config.json": "f3186777071146682ed0debc54b80ef604c42bf6728a1f7b6fbe7cf39b796c3b", "model.safetensors": "4787e37ad6b44b84e4403405e0f472b06b3bf681cbf7d162f16a9215870270a6", "processor_config.json": "b4a3962ea5f7ec39f40b5cf14e57ce99776c3dcce4756a110f7a169809e3a04c", "special_tokens_map.json": "abc97715b4b3b30eb65ea6895afd7b529c32bd10c28901f9ddae7edc39723b0f", "tokenizer.json": "9fb6333a7db8fe5da90728e741e4a3ee4ac2ae12c5dd4958cc6f31688787d3c2", "tokenizer_config.json": "2e36c9748a2081abb935b2e745ee22e82efa32589c2500df7e5bc0f93145cd77" } }