tarekziade HF Staff
Keep the real checkpoint's weight names, prefix and file format (tensors unchanged; reload verified bit-exact)
939ec1a verified | { | |
| "fixture": "huggingface-mic/conv-bert-base-mic", | |
| "key": "ccb41d493e60e4a925673f4784ded3645228beaec2f323c5b587c3f4a280baa4", | |
| "key_fields": { | |
| "architectures": [ | |
| "ConvBertModel" | |
| ], | |
| "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, | |
| "vocab_size": 30522, | |
| "hidden_size": 64, | |
| "num_hidden_layers": 2, | |
| "num_attention_heads": 4, | |
| "intermediate_size": 128, | |
| "hidden_act": "gelu", | |
| "hidden_dropout_prob": 0.1, | |
| "attention_probs_dropout_prob": 0.1, | |
| "max_position_embeddings": 512, | |
| "type_vocab_size": 2, | |
| "initializer_range": 0.02, | |
| "layer_norm_eps": 1e-12, | |
| "pad_token_id": 0, | |
| "bos_token_id": 0, | |
| "eos_token_id": 2, | |
| "embedding_size": 768, | |
| "head_ratio": 2, | |
| "conv_kernel_size": 9, | |
| "num_groups": 1, | |
| "classifier_dropout": null, | |
| "is_decoder": false, | |
| "add_cross_attention": false, | |
| "tie_word_embeddings": true, | |
| "model_type": "convbert", | |
| "output_attentions": false | |
| }, | |
| "covers": { | |
| "YituTech/conv-bert-base": "5cb451936b5c4a96562d8b146de85f64f9cf2c22" | |
| }, | |
| "built_from": { | |
| "repo": "YituTech/conv-bert-base", | |
| "revision": "5cb451936b5c4a96562d8b146de85f64f9cf2c22" | |
| }, | |
| "seed": 17, | |
| "dtype": "float32", | |
| "builder": "example/build_fixtures.py", | |
| "builder_revision": "d6c4e1368e76b3cb3fbd52f1d7813a1c321195c5", | |
| "recipe": { | |
| "seed": 17, | |
| "model_class": "ConvBertModel" | |
| }, | |
| "shrink": "rule", | |
| "model_class": "ConvBertModel", | |
| "parameters": 23951524, | |
| "missing_parameters": [], | |
| "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", | |
| "token_type_ids" | |
| ], | |
| "class": "ConvBertModel" | |
| }, | |
| "weights": "random", | |
| "gate": { | |
| "random": { | |
| "pass": true, | |
| "accepted_insensitive": {}, | |
| "threshold": 0.001, | |
| "deltas": { | |
| "input_ids:roll_batch": 2.6929197311401367, | |
| "input_ids:reverse": 2.810927391052246, | |
| "token_type_ids:all_ones": 2.8679213523864746, | |
| "attention_mask:all_ones": 0.0143318772315979 | |
| }, | |
| "weakest": "attention_mask:all_ones", | |
| "no_perturbation_defined": [], | |
| "not_a_forward_parameter": [] | |
| } | |
| }, | |
| "numeric_health": { | |
| "float32": { | |
| "outputs_finite": true, | |
| "finite": true | |
| }, | |
| "bfloat16": { | |
| "outputs_finite": true, | |
| "finite": true | |
| }, | |
| "float16": { | |
| "outputs_finite": true, | |
| "finite": true | |
| } | |
| }, | |
| "copied_unchanged": [ | |
| "vocab.txt" | |
| ], | |
| "environment": { | |
| "python": "3.12.12", | |
| "machine": "arm64", | |
| "torch": "2.14.1", | |
| "transformers": "5.22.0.dev0", | |
| "tokenizers": "0.23.3" | |
| }, | |
| "sha256": { | |
| "config.json": "360cc63ab105443d6cca46469d232f252932087c184a5b9f578f8723ab6918a9", | |
| "pytorch_model.bin": "48f3db6cd04690d9ee02963e30fa1708741bcae9e5180732731bca43a5929823", | |
| "vocab.txt": "ce0ab0a373c6ce1fe777eea87d927294379f2ce187e240f0a9e24d4e9ac22fa0" | |
| }, | |
| "serialization": { | |
| "file": "pytorch_model.bin", | |
| "reverted_conversions": [], | |
| "migrated_from": "model.safetensors" | |
| } | |
| } | |