Commit
Β·
011bc8a
1
Parent(s):
1d07414
README.md
Browse files
README.md
CHANGED
@@ -3,9 +3,9 @@ viewer: false
|
|
3 |
tags: [uv-script, classification, vllm, structured-outputs, gpu-required]
|
4 |
---
|
5 |
|
6 |
-
# Dataset Classification
|
7 |
|
8 |
-
|
9 |
|
10 |
## π Quick Start
|
11 |
|
@@ -22,21 +22,25 @@ That's it! No installation, no setup - just `uv run`.
|
|
22 |
|
23 |
## π Requirements
|
24 |
|
25 |
-
- **GPU Required**:
|
26 |
- Python 3.10+
|
27 |
- UV (will handle all dependencies automatically)
|
28 |
-
- vLLM >= 0.6.6
|
29 |
|
30 |
## π― Features
|
31 |
|
32 |
-
- **Guaranteed valid outputs** using
|
33 |
-
- **Zero-shot classification**
|
34 |
-
- **GPU-optimized**
|
35 |
-
- **Default model**: HuggingFaceTB/SmolLM3-3B (fast 3B model
|
36 |
- **Robust text handling** with preprocessing and validation
|
37 |
-
- **Three prompt styles** for different use cases
|
38 |
- **Automatic progress tracking** and detailed statistics
|
39 |
- **Direct Hub integration** - read and write datasets seamlessly
|
|
|
|
|
|
|
|
|
|
|
40 |
|
41 |
## π» Usage
|
42 |
|
@@ -62,20 +66,50 @@ uv run classify-dataset.py \
|
|
62 |
**Optional:**
|
63 |
|
64 |
- `--model`: Model to use (default: **`HuggingFaceTB/SmolLM3-3B`** - a fast 3B parameter model)
|
65 |
-
- `--
|
|
|
66 |
- `--split`: Dataset split to process (default: `train`)
|
67 |
- `--max-samples`: Limit samples for testing
|
|
|
|
|
68 |
- `--temperature`: Generation temperature (default: 0.1)
|
69 |
- `--guided-backend`: Backend for guided decoding (default: `outlines`)
|
70 |
- `--hf-token`: Hugging Face token (or use `HF_TOKEN` env var)
|
71 |
|
72 |
-
###
|
73 |
|
74 |
-
|
75 |
-
- **detailed**: Emphasizes exact category matching
|
76 |
-
- **reasoning**: Includes brief analysis before classification
|
77 |
|
78 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
79 |
|
80 |
## π Examples
|
81 |
|
@@ -96,8 +130,8 @@ uv run classify-dataset.py \
|
|
96 |
--input-dataset user/support-tickets \
|
97 |
--column content \
|
98 |
--labels "bug,feature_request,question,other" \
|
99 |
-
--
|
100 |
-
--
|
101 |
```
|
102 |
|
103 |
### News Categorization
|
@@ -111,11 +145,54 @@ uv run classify-dataset.py \
|
|
111 |
--model meta-llama/Llama-3.2-3B-Instruct
|
112 |
```
|
113 |
|
114 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
115 |
|
116 |
-
This
|
117 |
|
118 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
119 |
# Run on L4 GPU with vLLM image
|
120 |
hf jobs uv run \
|
121 |
--flavor l4x1 \
|
@@ -125,6 +202,7 @@ hf jobs uv run \
|
|
125 |
--column text \
|
126 |
--labels "positive,negative" \
|
127 |
--output-dataset user/imdb-classified
|
|
|
128 |
|
129 |
### GPU Flavors
|
130 |
- `t4-small`: Budget option for smaller models
|
@@ -135,6 +213,27 @@ hf jobs uv run \
|
|
135 |
|
136 |
## π§ Advanced Usage
|
137 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
138 |
### Using Different Models
|
139 |
|
140 |
By default, this script uses **HuggingFaceTB/SmolLM3-3B** - a fast, efficient 3B parameter model that's perfect for most classification tasks. You can easily use any other instruction-tuned model:
|
@@ -147,7 +246,7 @@ uv run classify-dataset.py \
|
|
147 |
--labels "contract,patent,brief,memo,other" \
|
148 |
--output-dataset user/legal-classified \
|
149 |
--model Qwen/Qwen2.5-7B-Instruct
|
150 |
-
|
151 |
|
152 |
### Large Datasets
|
153 |
|
@@ -166,15 +265,16 @@ uv run classify-dataset.py \
|
|
166 |
- **SmolLM3-3B (default)**: ~50-100 texts/second on A10
|
167 |
- **7B models**: ~20-50 texts/second on A10
|
168 |
- vLLM automatically optimizes batching for best throughput
|
|
|
169 |
|
170 |
## π€ How It Works
|
171 |
|
172 |
-
1. **vLLM**: Provides efficient GPU batch inference
|
173 |
-
2. **Guided Decoding**: Uses outlines to guarantee valid label outputs
|
174 |
3. **Structured Generation**: Constrains model outputs to exact label choices
|
175 |
4. **UV**: Handles all dependencies automatically
|
176 |
|
177 |
-
The script loads your dataset, preprocesses texts, classifies each one
|
178 |
|
179 |
## π Troubleshooting
|
180 |
|
@@ -212,43 +312,28 @@ If you see `ImportError: cannot import name 'GuidedDecodingParams'`:
|
|
212 |
- The script specifies the correct version in its dependencies
|
213 |
- UV should automatically install the correct version
|
214 |
|
215 |
-
## π¬ Advanced
|
216 |
|
217 |
-
For
|
218 |
|
219 |
-
|
220 |
-
|
221 |
-
|
222 |
-
|
223 |
-
uv run prepare_arxiv_2024.py
|
224 |
-
```
|
225 |
|
226 |
-
|
|
|
|
|
227 |
|
228 |
-
|
229 |
-
|
230 |
-
|
231 |
-
|
232 |
-
|
|
|
|
|
233 |
```
|
234 |
|
235 |
-
This script demonstrates:
|
236 |
-
|
237 |
-
- Using `run_uv_job()` from the Python API
|
238 |
-
- Classifying into modern ML trends (reasoning, agents, multimodal, robotics, etc.)
|
239 |
-
- Handling authentication and job monitoring
|
240 |
-
|
241 |
-
The classification categories include:
|
242 |
-
|
243 |
-
- `reasoning_systems`: Chain-of-thought, reasoning, problem solving
|
244 |
-
- `agents_autonomous`: Agents, tool use, autonomous systems
|
245 |
-
- `multimodal_models`: Vision-language, audio, multi-modal
|
246 |
-
- `robotics_embodied`: Robotics, embodied AI, manipulation
|
247 |
-
- `efficient_inference`: Quantization, distillation, edge deployment
|
248 |
-
- `alignment_safety`: RLHF, alignment, safety, interpretability
|
249 |
-
- `generative_models`: Diffusion, generation, synthesis
|
250 |
-
- `foundational_other`: Other foundational ML/AI research
|
251 |
-
|
252 |
## π License
|
253 |
|
254 |
This script is provided as-is for use with the UV Scripts organization.
|
|
|
3 |
tags: [uv-script, classification, vllm, structured-outputs, gpu-required]
|
4 |
---
|
5 |
|
6 |
+
# Dataset Classification Script
|
7 |
|
8 |
+
GPU-accelerated text classification for Hugging Face datasets with guaranteed valid outputs through structured generation.
|
9 |
|
10 |
## π Quick Start
|
11 |
|
|
|
22 |
|
23 |
## π Requirements
|
24 |
|
25 |
+
- **GPU Required**: Uses GPU-accelerated inference
|
26 |
- Python 3.10+
|
27 |
- UV (will handle all dependencies automatically)
|
28 |
+
- vLLM >= 0.6.6
|
29 |
|
30 |
## π― Features
|
31 |
|
32 |
+
- **Guaranteed valid outputs** using structured generation with guided decoding
|
33 |
+
- **Zero-shot classification** without training data required
|
34 |
+
- **GPU-optimized** for maximum throughput and efficiency
|
35 |
+
- **Default model**: HuggingFaceTB/SmolLM3-3B (fast 3B model with thinking capabilities)
|
36 |
- **Robust text handling** with preprocessing and validation
|
|
|
37 |
- **Automatic progress tracking** and detailed statistics
|
38 |
- **Direct Hub integration** - read and write datasets seamlessly
|
39 |
+
- **Label descriptions** support for providing context to improve accuracy
|
40 |
+
- **Reasoning mode** for interpretable classifications with thinking traces
|
41 |
+
- **JSON output parsing** for reliable extraction from reasoning mode
|
42 |
+
- **Optimized batching** with vLLM's automatic batch processing
|
43 |
+
- **Multiple guided backends** - supports outlines, xgrammar, and more
|
44 |
|
45 |
## π» Usage
|
46 |
|
|
|
66 |
**Optional:**
|
67 |
|
68 |
- `--model`: Model to use (default: **`HuggingFaceTB/SmolLM3-3B`** - a fast 3B parameter model)
|
69 |
+
- `--label-descriptions`: Provide descriptions for each label to improve classification accuracy
|
70 |
+
- `--enable-reasoning`: Enable reasoning mode with thinking traces (adds reasoning column)
|
71 |
- `--split`: Dataset split to process (default: `train`)
|
72 |
- `--max-samples`: Limit samples for testing
|
73 |
+
- `--shuffle`: Shuffle dataset before selecting samples (useful for random sampling)
|
74 |
+
- `--shuffle-seed`: Random seed for shuffling (default: 42)
|
75 |
- `--temperature`: Generation temperature (default: 0.1)
|
76 |
- `--guided-backend`: Backend for guided decoding (default: `outlines`)
|
77 |
- `--hf-token`: Hugging Face token (or use `HF_TOKEN` env var)
|
78 |
|
79 |
+
### Label Descriptions
|
80 |
|
81 |
+
Provide context for your labels to improve classification accuracy:
|
|
|
|
|
82 |
|
83 |
+
```bash
|
84 |
+
uv run classify-dataset.py \
|
85 |
+
--input-dataset user/support-tickets \
|
86 |
+
--column content \
|
87 |
+
--labels "bug,feature,question,other" \
|
88 |
+
--label-descriptions "bug:something is broken,feature:request for new functionality,question:asking for help,other:anything else" \
|
89 |
+
--output-dataset user/tickets-classified
|
90 |
+
```
|
91 |
+
|
92 |
+
The model uses these descriptions to better understand what each label represents, leading to more accurate classifications.
|
93 |
+
|
94 |
+
### Reasoning Mode
|
95 |
+
|
96 |
+
Enable thinking traces for interpretable classifications:
|
97 |
+
|
98 |
+
```bash
|
99 |
+
uv run classify-dataset.py \
|
100 |
+
--input-dataset stanfordnlp/imdb \
|
101 |
+
--column text \
|
102 |
+
--labels "positive,negative,neutral" \
|
103 |
+
--enable-reasoning \
|
104 |
+
--output-dataset user/imdb-with-reasoning
|
105 |
+
```
|
106 |
+
|
107 |
+
When `--enable-reasoning` is used:
|
108 |
+
- The model generates step-by-step reasoning using SmolLM3's thinking capabilities
|
109 |
+
- Output includes three columns: `classification`, `reasoning`, and `parsing_success`
|
110 |
+
- Final answer must be in JSON format: `{"label": "chosen_label"}`
|
111 |
+
- Useful for understanding complex classification decisions
|
112 |
+
- Trade-off: Slower but more interpretable
|
113 |
|
114 |
## π Examples
|
115 |
|
|
|
130 |
--input-dataset user/support-tickets \
|
131 |
--column content \
|
132 |
--labels "bug,feature_request,question,other" \
|
133 |
+
--label-descriptions "bug:code or product not working as expected,feature_request:asking for new functionality,question:seeking help or clarification,other:general comments or feedback" \
|
134 |
+
--output-dataset user/tickets-classified
|
135 |
```
|
136 |
|
137 |
### News Categorization
|
|
|
145 |
--model meta-llama/Llama-3.2-3B-Instruct
|
146 |
```
|
147 |
|
148 |
+
### Complex Classification with Reasoning
|
149 |
+
|
150 |
+
```bash
|
151 |
+
uv run classify-dataset.py \
|
152 |
+
--input-dataset user/customer-feedback \
|
153 |
+
--column text \
|
154 |
+
--labels "very_positive,positive,neutral,negative,very_negative" \
|
155 |
+
--label-descriptions "very_positive:extremely satisfied,positive:generally satisfied,neutral:mixed feelings,negative:dissatisfied,very_negative:extremely dissatisfied" \
|
156 |
+
--enable-reasoning \
|
157 |
+
--output-dataset user/feedback-analyzed
|
158 |
+
```
|
159 |
|
160 |
+
This combines label descriptions with reasoning mode for maximum interpretability.
|
161 |
|
162 |
+
### ArXiv ML Research Classification
|
163 |
+
|
164 |
+
Classify academic papers into machine learning research areas:
|
165 |
+
|
166 |
+
```bash
|
167 |
+
# Fast classification with random sampling
|
168 |
+
uv run classify-dataset.py \
|
169 |
+
--input-dataset librarian-bots/arxiv-metadata-snapshot \
|
170 |
+
--column abstract \
|
171 |
+
--labels "llm,computer_vision,reinforcement_learning,optimization,theory,other" \
|
172 |
+
--label-descriptions "llm:language models and NLP,computer_vision:image and video processing,reinforcement_learning:RL and decision making,optimization:training and efficiency,theory:theoretical ML foundations,other:other ML topics" \
|
173 |
+
--output-dataset user/arxiv-ml-classified \
|
174 |
+
--split "train[:10000]" \
|
175 |
+
--max-samples 100 \
|
176 |
+
--shuffle
|
177 |
+
|
178 |
+
# With reasoning for nuanced classification
|
179 |
+
uv run classify-dataset.py \
|
180 |
+
--input-dataset librarian-bots/arxiv-metadata-snapshot \
|
181 |
+
--column abstract \
|
182 |
+
--labels "multimodal,agents,reasoning,safety,efficiency" \
|
183 |
+
--label-descriptions "multimodal:vision-language and cross-modal models,agents:autonomous agents and tool use,reasoning:reasoning and planning systems,safety:alignment and safety research,efficiency:model optimization and deployment" \
|
184 |
+
--enable-reasoning \
|
185 |
+
--output-dataset user/arxiv-frontier-research \
|
186 |
+
--split "train[:1000]" \
|
187 |
+
--max-samples 50
|
188 |
+
```
|
189 |
+
|
190 |
+
The reasoning mode is particularly valuable for academic abstracts where papers often span multiple topics and require careful analysis to determine the primary focus.
|
191 |
+
|
192 |
+
## π Running on HF Jobs
|
193 |
+
|
194 |
+
Optimized for [Hugging Face Jobs](https://huggingface.co/docs/hub/spaces-gpu-jobs) (requires Pro subscription or Team/Enterprise organization):
|
195 |
+
```bash
|
196 |
# Run on L4 GPU with vLLM image
|
197 |
hf jobs uv run \
|
198 |
--flavor l4x1 \
|
|
|
202 |
--column text \
|
203 |
--labels "positive,negative" \
|
204 |
--output-dataset user/imdb-classified
|
205 |
+
```
|
206 |
|
207 |
### GPU Flavors
|
208 |
- `t4-small`: Budget option for smaller models
|
|
|
213 |
|
214 |
## π§ Advanced Usage
|
215 |
|
216 |
+
### Random Sampling
|
217 |
+
|
218 |
+
When working with ordered datasets, use `--shuffle` with `--max-samples` to get a representative sample:
|
219 |
+
|
220 |
+
```bash
|
221 |
+
# Get 50 random reviews instead of the first 50
|
222 |
+
uv run classify-dataset.py \
|
223 |
+
--input-dataset stanfordnlp/imdb \
|
224 |
+
--column text \
|
225 |
+
--labels "positive,negative" \
|
226 |
+
--output-dataset user/imdb-sample \
|
227 |
+
--max-samples 50 \
|
228 |
+
--shuffle \
|
229 |
+
--shuffle-seed 123 # For reproducibility
|
230 |
+
```
|
231 |
+
|
232 |
+
This is especially important for:
|
233 |
+
- Chronologically ordered datasets (news, papers, social media)
|
234 |
+
- Pre-sorted datasets (by rating, category, etc.)
|
235 |
+
- Testing on diverse samples before processing the full dataset
|
236 |
+
|
237 |
### Using Different Models
|
238 |
|
239 |
By default, this script uses **HuggingFaceTB/SmolLM3-3B** - a fast, efficient 3B parameter model that's perfect for most classification tasks. You can easily use any other instruction-tuned model:
|
|
|
246 |
--labels "contract,patent,brief,memo,other" \
|
247 |
--output-dataset user/legal-classified \
|
248 |
--model Qwen/Qwen2.5-7B-Instruct
|
249 |
+
```
|
250 |
|
251 |
### Large Datasets
|
252 |
|
|
|
265 |
- **SmolLM3-3B (default)**: ~50-100 texts/second on A10
|
266 |
- **7B models**: ~20-50 texts/second on A10
|
267 |
- vLLM automatically optimizes batching for best throughput
|
268 |
+
- Performance scales with GPU memory and compute capability
|
269 |
|
270 |
## π€ How It Works
|
271 |
|
272 |
+
1. **vLLM**: Provides efficient GPU batch inference with automatic batching
|
273 |
+
2. **Guided Decoding**: Uses outlines backend to guarantee valid label outputs
|
274 |
3. **Structured Generation**: Constrains model outputs to exact label choices
|
275 |
4. **UV**: Handles all dependencies automatically
|
276 |
|
277 |
+
The script loads your dataset, preprocesses texts, classifies each one with guaranteed valid outputs, then saves the results as a new column in the output dataset.
|
278 |
|
279 |
## π Troubleshooting
|
280 |
|
|
|
312 |
- The script specifies the correct version in its dependencies
|
313 |
- UV should automatically install the correct version
|
314 |
|
315 |
+
## π¬ Advanced Workflows
|
316 |
|
317 |
+
For complex real-world workflows that integrate UV scripts with the Python HF Jobs API, see the [ArXiv ML Trends example](examples/arxiv-workflow/). This demonstrates:
|
318 |
|
319 |
+
- **Multi-stage pipelines**: Data preparation β GPU classification β Analysis
|
320 |
+
- **Python API orchestration**: Using `run_uv_job()` to manage GPU jobs programmatically
|
321 |
+
- **Production patterns**: Error handling, parallel execution, and incremental updates
|
322 |
+
- **Cost optimization**: Choosing appropriate compute resources for each task
|
|
|
|
|
323 |
|
324 |
+
```python
|
325 |
+
# Example: Submit a classification job via Python API
|
326 |
+
from huggingface_hub import run_uv_job
|
327 |
|
328 |
+
job = run_uv_job(
|
329 |
+
script="https://huggingface.co/datasets/uv-scripts/classification/raw/main/classify-dataset.py",
|
330 |
+
args=["--input-dataset", "my/dataset", "--labels", "A,B,C"],
|
331 |
+
flavor="l4x1",
|
332 |
+
image="vllm/vllm-openai:latest"
|
333 |
+
)
|
334 |
+
result = job.wait()
|
335 |
```
|
336 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
337 |
## π License
|
338 |
|
339 |
This script is provided as-is for use with the UV Scripts organization.
|