Upload MLX packaging for OpenMed-NER-ChemicalDetect-SuperClinical-434M-mlx

README.md

@@ -0,0 +1,79 @@
+---
+license: apache-2.0
+base_model: OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M
+pipeline_tag: token-classification
+library_name: openmed
+tags:
+  - openmed
+  - mlx
+  - apple-silicon
+  - token-classification
+  - pii
+  - de-identification
+  - medical
+  - clinical
+---
+
+# OpenMed-NER-ChemicalDetect-SuperClinical-434M for OpenMed MLX
+
+This repository contains an OpenMed MLX conversion of [`OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M`](https://huggingface.co/OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M) for Apple Silicon inference with [OpenMed](https://github.com/maziyarpanahi/openmed).
+
+Artifact metadata:
+
+- OpenMed MLX task: `token-classification`
+- OpenMed MLX family: `deberta-v2`
+- Weight format: `safetensors`
+- Runtime API: `OpenMed MLX token-classification backend`
+
+[OpenMed](https://github.com/maziyarpanahi/openmed) is the main product experience:
+
+- Install the Python package with `pip install openmed`
+- Enable Apple Silicon acceleration with `pip install "openmed[mlx]"`
+- Load this MLX model directly from the Hub or from a local snapshot
+- For Apple apps, use OpenMedKit from the same GitHub repository with a compatible CoreML bundle
+
+This MLX repo is meant to pair with:
+
+- OpenMed GitHub: [https://github.com/maziyarpanahi/openmed](https://github.com/maziyarpanahi/openmed)
+- OpenMed website: [https://openmed.life](https://openmed.life)
+- Source checkpoint: [`OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M`](https://huggingface.co/OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M)
+
+## Quick Start
+
+```bash
+pip install openmed
+pip install "openmed[mlx]"
+```
+
+```python
+from openmed import analyze_text
+from openmed.core.config import OpenMedConfig
+
+result = analyze_text(
+    "Patient John Doe, DOB 1990-05-15, SSN 123-45-6789",
+    model_name="OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M",
+    config=OpenMedConfig(backend="mlx"),
+)
+
+for entity in result.entities:
+    print(entity.label, entity.text, round(entity.confidence, 4))
+```
+
+## Swift and Apple Apps
+
+Use Swift with OpenMedKit, not with MLX weight files directly.
+
+1. Open Xcode and go to File > Add Package Dependencies.
+2. Paste the OpenMed repository URL: `https://github.com/maziyarpanahi/openmed`
+3. Choose the package product OpenMedKit from the repository.
+4. Add a compatible CoreML model bundle plus `id2label.json` to your app target.
+
+This MLX model is for Python services on Apple Silicon, local MLX inference on macOS, and Hub-hosted model distribution. If a given environment cannot write `weights.safetensors`, OpenMed falls back to `weights.npz` so the model remains usable.
+
+## Credits
+
+- Base checkpoint: [`OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M`](https://huggingface.co/OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M)
+- OpenMed GitHub: [https://github.com/maziyarpanahi/openmed](https://github.com/maziyarpanahi/openmed)
+- OpenMed website: [https://openmed.life](https://openmed.life)
+- MLX conversion and runtime support: OpenMed
+- Swift runtime for Apple apps: OpenMedKit from the OpenMed repository

config.json

@@ -0,0 +1,61 @@
+{
+  "transformers_version": "5.5.0",
+  "architectures": [
+    "DebertaV2ForTokenClassification"
+  ],
+  "output_hidden_states": false,
+  "return_dict": true,
+  "dtype": "bfloat16",
+  "chunk_size_feed_forward": 0,
+  "is_encoder_decoder": false,
+  "id2label": {
+    "0": "O",
+    "1": "B-CHEM",
+    "2": "I-CHEM"
+  },
+  "label2id": {
+    "B-CHEM": 1,
+    "I-CHEM": 2,
+    "O": 0
+  },
+  "problem_type": null,
+  "vocab_size": 128100,
+  "hidden_size": 1024,
+  "num_hidden_layers": 24,
+  "num_attention_heads": 16,
+  "intermediate_size": 4096,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.2,
+  "attention_probs_dropout_prob": 0.2,
+  "max_position_embeddings": 512,
+  "type_vocab_size": 0,
+  "initializer_range": 0.02,
+  "layer_norm_eps": 1e-07,
+  "relative_attention": true,
+  "max_relative_positions": -1,
+  "pad_token_id": 0,
+  "bos_token_id": null,
+  "eos_token_id": null,
+  "position_biased_input": false,
+  "pos_att_type": [
+    "p2c",
+    "c2p"
+  ],
+  "pooler_dropout": 0,
+  "pooler_hidden_act": "gelu",
+  "legacy": true,
+  "tie_word_embeddings": true,
+  "pooler_hidden_size": 1024,
+  "_name_or_path": "OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M",
+  "model_type": "deberta-v2",
+  "norm_rel_ebd": "layer_norm",
+  "position_buckets": 256,
+  "share_att_key": true,
+  "output_attentions": false,
+  "_mlx_task": "token-classification",
+  "_mlx_family": "deberta-v2",
+  "_mlx_position_offset": 0,
+  "_mlx_model_type": "deberta-v2",
+  "num_labels": 3,
+  "_mlx_weights_format": "safetensors"
+}

id2label.json

@@ -0,0 +1,5 @@
+{
+  "0": "O",
+  "1": "B-CHEM",
+  "2": "I-CHEM"
+}

openmed-mlx.json

@@ -0,0 +1,26 @@
+{
+  "format": "openmed-mlx",
+  "format_version": 2,
+  "task": "token-classification",
+  "family": "deberta-v2",
+  "source_model_id": "OpenMed/OpenMed-NER-ChemicalDetect-SuperClinical-434M",
+  "config_path": "config.json",
+  "label_map_path": "id2label.json",
+  "preferred_weights": "weights.safetensors",
+  "fallback_weights": [
+    "weights.npz"
+  ],
+  "available_weights": [
+    "weights.safetensors"
+  ],
+  "weights_format": "safetensors",
+  "quantization": null,
+  "max_sequence_length": 512,
+  "tokenizer": {
+    "path": ".",
+    "files": [
+      "tokenizer.json",
+      "tokenizer_config.json"
+    ]
+  }
+}

tokenizer_config.json

@@ -0,0 +1,20 @@
+{
+  "add_prefix_space": true,
+  "backend": "tokenizers",
+  "bos_token": "[CLS]",
+  "clean_up_tokenization_spaces": false,
+  "cls_token": "[CLS]",
+  "do_lower_case": false,
+  "eos_token": "[SEP]",
+  "is_local": false,
+  "mask_token": "[MASK]",
+  "model_max_length": 1000000000000000019884624838656,
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "sp_model_kwargs": {},
+  "split_by_punct": false,
+  "tokenizer_class": "DebertaV2Tokenizer",
+  "unk_id": 3,
+  "unk_token": "[UNK]",
+  "vocab_type": "spm"
+}

weights.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4957f9124afdd133e7d1a3b693e84243a7d43c4b33e90deaecce2bbbf24c52d0
+size 1736107082