{ "fixture": "huggingface-mic/videomae-base-mic", "key": "4644e14e0e13445fdfb7644501b0fc57ca5fd2efe614d1cb23ac7326669e6e40", "key_fields": { "architectures": [ "VideoMAEForPreTraining" ], "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, "image_size": 224, "patch_size": 16, "num_channels": 3, "num_frames": 16, "tubelet_size": 2, "hidden_size": 64, "num_hidden_layers": 2, "num_attention_heads": 4, "intermediate_size": 128, "hidden_act": "gelu", "hidden_dropout_prob": 0.0, "attention_probs_dropout_prob": 0.0, "initializer_range": 0.02, "layer_norm_eps": 1e-12, "qv_bias": true, "use_mean_pooling": false, "decoder_num_attention_heads": 6, "decoder_hidden_size": 384, "decoder_num_hidden_layers": 4, "decoder_intermediate_size": 1536, "norm_pix_loss": true, "model_type": "videomae", "output_attentions": false }, "covers": { "MCG-NJU/videomae-base": "dc740ceda42fce44faed2ea03c6d447db72f6af9" }, "built_from": { "repo": "MCG-NJU/videomae-base", "revision": "dc740ceda42fce44faed2ea03c6d447db72f6af9" }, "seed": 17, "dtype": "float32", "builder": "example/build_fixtures.py", "builder_revision": "1c134326190e21d73b5a77df351cc521b9970f3b", "recipe": { "seed": 17, "model_class": "VideoMAEForPreTraining" }, "shrink": "rule", "model_class": "VideoMAEForPreTraining", "parameters": 7878720, "missing_parameters": [], "serialization": { "file": "model.safetensors", "source_file": "model.safetensors", "reverted_conversions": [ "WeightRenaming: ['attention\\\\.attention\\\\.q_bias'] -> ['attention.attention.query.bias']", "WeightRenaming: ['attention\\\\.attention\\\\.v_bias'] -> ['attention.attention.value.bias']" ] }, "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": [ "bool_masked_pos", "pixel_values" ], "class": "VideoMAEForPreTraining" }, "weights": "random", "gate": { "random": { "pass": true, "accepted_insensitive": {}, "threshold": 0.001, "deltas": { "pixel_values:roll_batch": 0.0023539960384368896, "pixel_values:swap_channels": 0.0057561397552490234, "pixel_values:reverse": 0.007363468408584595, "pixel_values:halve": 0.011716663837432861 }, "weakest": "pixel_values:roll_batch", "no_perturbation_defined": [ "bool_masked_pos" ], "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": [ "preprocessor_config.json" ], "environment": { "python": "3.11.18", "machine": "x86_64", "torch": "2.15.0+cpu", "transformers": "5.23.0.dev0", "tokenizers": "0.23.3" }, "sha256": { "config.json": "51a1a629e587f60d4eba02b53120cca593502d8f6e89f583424794c7aadfc6d2", "model.safetensors": "047b1da6b377770accdaefbb9aa9ee5a13c0b37cdb4d09f087655ffed066615f", "preprocessor_config.json": "c3aa722f22a7ff0d234d407862025fe47672f74f46f92c91f757f7a7354107c2" } }