tarekziade HF Staff
Publish MIC usage, slim fixture and reviewed Transformers expectations
0ad90af verified | { | |
| "fixture": "huggingface-mic/tiny_starcoder_py-mic", | |
| "key": "0852b1b7488daf526180d9c8c3361de942729124e3b94beb0339fa2d46cf47e0", | |
| "key_fields": { | |
| "architectures": [ | |
| "GPTBigCodeForCausalLM" | |
| ], | |
| "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": 49152, | |
| "n_positions": 8192, | |
| "n_embd": 64, | |
| "n_layer": 2, | |
| "n_head": 4, | |
| "n_inner": 128, | |
| "activation_function": "gelu_pytorch_tanh", | |
| "resid_pdrop": 0.1, | |
| "embd_pdrop": 0.1, | |
| "attn_pdrop": 0.1, | |
| "layer_norm_epsilon": 1e-05, | |
| "initializer_range": 0.02, | |
| "scale_attn_weights": true, | |
| "use_cache": true, | |
| "bos_token_id": 0, | |
| "eos_token_id": 0, | |
| "pad_token_id": null, | |
| "attention_softmax_in_fp32": true, | |
| "scale_attention_softmax_in_fp32": true, | |
| "multi_query": true, | |
| "add_cross_attention": false, | |
| "tie_word_embeddings": true, | |
| "num_key_value_heads": 2, | |
| "inference_runner": 0, | |
| "max_batch_size": null, | |
| "max_sequence_length": null, | |
| "model_type": "gpt_bigcode", | |
| "pad_key_length": true, | |
| "pre_allocate_kv_cache": false, | |
| "summary_activation": null, | |
| "summary_first_dropout": 0.1, | |
| "summary_proj_to_labels": true, | |
| "summary_type": "cls_index", | |
| "summary_use_proj": true, | |
| "validate_runner_input": true, | |
| "output_attentions": false | |
| }, | |
| "covers": { | |
| "bigcode/tiny_starcoder_py": "8547527bef0bc927268c1653cce6948c5c242dd1" | |
| }, | |
| "built_from": { | |
| "repo": "bigcode/tiny_starcoder_py", | |
| "revision": "8547527bef0bc927268c1653cce6948c5c242dd1" | |
| }, | |
| "seed": 17, | |
| "dtype": "float32", | |
| "builder": "example/build_fixtures.py", | |
| "builder_revision": "c0a9ac9ce31d96d81a7ed980688dcd4599b1e6bd", | |
| "recipe": { | |
| "seed": 17, | |
| "model_class": "GPTBigCodeForCausalLM" | |
| }, | |
| "shrink": "rule", | |
| "model_class": "GPTBigCodeForCausalLM", | |
| "parameters": 3724608, | |
| "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", | |
| "labels" | |
| ], | |
| "class": "GPTBigCodeForCausalLM" | |
| }, | |
| "weights": "random", | |
| "gate": { | |
| "random": { | |
| "pass": true, | |
| "accepted_insensitive": {}, | |
| "threshold": 0.001, | |
| "deltas": { | |
| "input_ids:roll_batch": 1.1869170665740967, | |
| "input_ids:reverse": 1.1730904579162598, | |
| "attention_mask:all_ones": 0.49951428174972534 | |
| }, | |
| "weakest": "attention_mask:all_ones", | |
| "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": [ | |
| "generation_config.json", | |
| "merges.txt", | |
| "special_tokens_map.json", | |
| "tokenizer.json", | |
| "tokenizer_config.json", | |
| "vocab.json" | |
| ], | |
| "environment": { | |
| "python": "3.12.12", | |
| "machine": "arm64", | |
| "torch": "2.14.1", | |
| "transformers": "5.22.0.dev0", | |
| "tokenizers": "0.23.3" | |
| }, | |
| "sha256": { | |
| "config.json": "e71a18da111f128967e7d8cc89622e71dafc295a27b8181170d11423ad420ea0", | |
| "generation_config.json": "368f15f9c335d48e1a585440037c77022af6b867303dc8fe9ee01e8fc86c1617", | |
| "merges.txt": "74b0a4bc1a97ebc1d227f69231b18574374ba052fb945b0fa0aa91d3c32504a2", | |
| "model.safetensors": "1cb730824fb0a3d7a02141afb42177445e7d8522e8cde09bbfb3c174d5ca2ed0", | |
| "special_tokens_map.json": "0823292e24ea07b89317e9ede9d08da2a1b6c014290c06908a7ad04f1efd6719", | |
| "tokenizer.json": "42b5a37ba11199f024f2b8873e1ecba98da33166e16f700bf7cb2304b0a5583f", | |
| "tokenizer_config.json": "95684c52ad9a970dbbb17576ee2237cb62902c1eff6804c7c91a4d6219a4a6d7", | |
| "vocab.json": "20175afb9f164fad4829aca2279f8df7eeff1e2e3f671378aaa287a740aff09f" | |
| } | |
| } | |