tarekziade HF Staff
Keep the real checkpoint's weight names, prefix and file format (tensors unchanged; reload verified bit-exact)
0b11088 verified | { | |
| "fixture": "huggingface-mic/fnet-base-mic", | |
| "key": "0f5decac063fd1ac4cef0d20921e7f5e8bbdacb861ca3bb965c132c2f3022b05", | |
| "key_fields": { | |
| "architectures": [ | |
| "FNetForPreTraining" | |
| ], | |
| "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": 32000, | |
| "hidden_size": 64, | |
| "num_hidden_layers": 2, | |
| "intermediate_size": 128, | |
| "hidden_act": "gelu_new", | |
| "hidden_dropout_prob": 0.1, | |
| "max_position_embeddings": 512, | |
| "type_vocab_size": 4, | |
| "initializer_range": 0.02, | |
| "layer_norm_eps": 1e-12, | |
| "use_tpu_fourier_optimizations": false, | |
| "tpu_short_seq_length": 512, | |
| "pad_token_id": 3, | |
| "bos_token_id": 1, | |
| "eos_token_id": 2, | |
| "tie_word_embeddings": true, | |
| "actual_seq_length": 512, | |
| "model_type": "fnet", | |
| "use_latest": false, | |
| "use_fft": true, | |
| "output_attentions": false | |
| }, | |
| "covers": { | |
| "google/fnet-base": "d89b6fad3cf5384848b783dc480f9685f49d008c" | |
| }, | |
| "built_from": { | |
| "repo": "google/fnet-base", | |
| "revision": "d89b6fad3cf5384848b783dc480f9685f49d008c" | |
| }, | |
| "seed": 17, | |
| "dtype": "float32", | |
| "builder": "example/build_fixtures.py", | |
| "builder_revision": "e60cfd37adc70f455a7adcba3317dff71b612f80", | |
| "recipe": { | |
| "seed": 17, | |
| "model_class": "FNetForPreTraining", | |
| "accept_unhealthy": { | |
| "bfloat16": "FNet's Fourier mixing rejects half precision on CPU: 'mixed dtype (CPU): expect parameter to have scalar type of Float'", | |
| "float16": "FNet's Fourier mixing rejects half precision on CPU: 'mixed dtype (CPU): expect parameter to have scalar type of Float'" | |
| } | |
| }, | |
| "shrink": "rule", | |
| "model_class": "FNetForPreTraining", | |
| "parameters": 2159554, | |
| "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": [ | |
| "input_ids", | |
| "labels", | |
| "next_sentence_label", | |
| "token_type_ids" | |
| ], | |
| "class": "FNetForPreTraining" | |
| }, | |
| "weights": "random", | |
| "gate": { | |
| "random": { | |
| "pass": true, | |
| "accepted_insensitive": {}, | |
| "threshold": 0.001, | |
| "deltas": { | |
| "input_ids:roll_batch": 0.9135565757751465, | |
| "input_ids:reverse": 0.8673748970031738, | |
| "token_type_ids:all_ones": 0.3061041235923767 | |
| }, | |
| "weakest": "token_type_ids:all_ones", | |
| "no_perturbation_defined": [ | |
| "next_sentence_label" | |
| ], | |
| "not_a_forward_parameter": [] | |
| } | |
| }, | |
| "numeric_health": { | |
| "float32": { | |
| "outputs_finite": true, | |
| "loss_finite": true, | |
| "gradients_finite": true, | |
| "finite": true | |
| }, | |
| "bfloat16": { | |
| "finite": false, | |
| "error": "RuntimeError: mixed dtype (CPU): expect parameter to have scalar type of Float" | |
| }, | |
| "float16": { | |
| "finite": false, | |
| "error": "RuntimeError: mixed dtype (CPU): expect parameter to have scalar type of Float" | |
| } | |
| }, | |
| "copied_unchanged": [ | |
| "special_tokens_map.json", | |
| "spiece.model", | |
| "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": { | |
| "config.json": "43bc6cc250df5477df54dc4871f051554cc10f834a9d06fe6a336fc5939f1e97", | |
| "pytorch_model.bin": "2513e2b5fc5ad6facf91761c9a7fc0b5ca091479351de59327f8a28534759319", | |
| "special_tokens_map.json": "447e507ea70c8b82c296b93486f9ab6a3090788e1e4516eed7f1c576c11ed149", | |
| "spiece.model": "0f848afb4cb35389f15819aad6b9b4bb65d8b6cec9613b6145bea94b4a045ab6", | |
| "tokenizer.json": "e19ac913a3dba26fc3aca52d60ae2a8e0b6a577b1c247de2f7cd6d192fa2dd05", | |
| "tokenizer_config.json": "5c2802f6f0ad6eadcb966a643b4a4c5cf3b5da6c84cbec71209eceb21538fae7" | |
| }, | |
| "serialization": { | |
| "file": "pytorch_model.bin", | |
| "reverted_conversions": [], | |
| "migrated_from": "model.safetensors" | |
| } | |
| } | |