Token Classification
MLX
eurobert
html
content-extraction
boilerplate-removal
web-scraping
encoder
custom-code
custom_code
Instructions to use Mike0021/pulpie-orange-small-mlx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use Mike0021/pulpie-orange-small-mlx with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir pulpie-orange-small-mlx Mike0021/pulpie-orange-small-mlx
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
| { | |
| "compute": { | |
| "environment": "Linux x86_64 CPU with mlx[cpu]; no paid cloud Mac used.", | |
| "estimated_incremental_cost_usd": 0.0 | |
| }, | |
| "end_to_end_extraction": { | |
| "chunk_lengths": [ | |
| 50 | |
| ], | |
| "input_html": "<html><body><article><h1>Apple MLX conversion</h1><p>This article explains how to convert a EuroBERT content extraction model to MLX format.</p></article></body></html>", | |
| "num_blocks": 2, | |
| "variants": { | |
| "4bit": { | |
| "html": "<html><body><article><h1 _item_id=\"1\">Apple MLX conversion</h1><p _item_id=\"2\">This article explains how to convert a EuroBERT content extraction model to MLX format.</p></article></body></html>", | |
| "labels": { | |
| "1": "main", | |
| "2": "main" | |
| }, | |
| "latency_ms": 77852.023473999, | |
| "markdown": "<html><body><article><h1 _item_id=\"1\">Apple MLX conversion</h1><p _item_id=\"2\">This article explains how to convert a EuroBERT content extraction model to MLX format.</p></article></body></html>", | |
| "non_empty": true, | |
| "predictions": [ | |
| 1, | |
| 1 | |
| ] | |
| }, | |
| "8bit": { | |
| "html": "<html><body><article><h1 _item_id=\"1\">Apple MLX conversion</h1><p _item_id=\"2\">This article explains how to convert a EuroBERT content extraction model to MLX format.</p></article></body></html>", | |
| "labels": { | |
| "1": "main", | |
| "2": "main" | |
| }, | |
| "latency_ms": 78333.6143399938, | |
| "markdown": "<html><body><article><h1 _item_id=\"1\">Apple MLX conversion</h1><p _item_id=\"2\">This article explains how to convert a EuroBERT content extraction model to MLX format.</p></article></body></html>", | |
| "non_empty": true, | |
| "predictions": [ | |
| 1, | |
| 1 | |
| ] | |
| }, | |
| "bf16": { | |
| "html": "<html><body><article><h1 _item_id=\"1\">Apple MLX conversion</h1><p _item_id=\"2\">This article explains how to convert a EuroBERT content extraction model to MLX format.</p></article></body></html>", | |
| "labels": { | |
| "1": "main", | |
| "2": "main" | |
| }, | |
| "latency_ms": 5469.0545329940505, | |
| "markdown": "<html><body><article><h1 _item_id=\"1\">Apple MLX conversion</h1><p _item_id=\"2\">This article explains how to convert a EuroBERT content extraction model to MLX format.</p></article></body></html>", | |
| "non_empty": true, | |
| "predictions": [ | |
| 1, | |
| 1 | |
| ] | |
| } | |
| } | |
| }, | |
| "load_checks": { | |
| "4bit": { | |
| "loaded": true, | |
| "logits_dtype": "mlx.core.bfloat16", | |
| "logits_shape": [ | |
| 1, | |
| 3, | |
| 2 | |
| ] | |
| }, | |
| "8bit": { | |
| "loaded": true, | |
| "logits_dtype": "mlx.core.bfloat16", | |
| "logits_shape": [ | |
| 1, | |
| 3, | |
| 2 | |
| ] | |
| }, | |
| "bf16": { | |
| "loaded": true, | |
| "logits_dtype": "mlx.core.bfloat16", | |
| "logits_shape": [ | |
| 1, | |
| 3, | |
| 2 | |
| ] | |
| } | |
| }, | |
| "model_dir": "hf_out", | |
| "numerical_accuracy": { | |
| "test_inputs": [ | |
| "A", | |
| "B", | |
| "C" | |
| ], | |
| "token_shape": [ | |
| 3, | |
| 2 | |
| ], | |
| "torch_reference": { | |
| "attention": "eager", | |
| "dtype": "float32", | |
| "latency_ms": 50.09009799687192 | |
| }, | |
| "variants": { | |
| "4bit": { | |
| "latency_ms": 9323.177823971491, | |
| "max_abs_diff": 2.2551989555358887, | |
| "mean_abs_diff": 1.1897996664047241 | |
| }, | |
| "8bit": { | |
| "latency_ms": 9380.578643002082, | |
| "max_abs_diff": 1.2797489166259766, | |
| "mean_abs_diff": 0.542361319065094 | |
| }, | |
| "bf16": { | |
| "latency_ms": 716.8781639775261, | |
| "max_abs_diff": 0.04523277282714844, | |
| "mean_abs_diff": 0.019181733950972557 | |
| } | |
| } | |
| }, | |
| "source_model": "feyninc/pulpie-orange-small", | |
| "variants": [ | |
| "bf16", | |
| "8bit", | |
| "4bit" | |
| ] | |
| } | |