Text Classification
Transformers
Safetensors
modernbert
routing
llm-routing
prompt-classification
multilingual
text-embeddings-inference
Instructions to use scombat/s1d1-router-6src-compact with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use scombat/s1d1-router-6src-compact with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="scombat/s1d1-router-6src-compact")# pip install -U transformers accelerate # Load model directly from transformers import AutoTokenizer, AutoModelForSequenceClassification tokenizer = AutoTokenizer.from_pretrained("scombat/s1d1-router-6src-compact") model = AutoModelForSequenceClassification.from_pretrained("scombat/s1d1-router-6src-compact", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Card update
Browse files
README.md
CHANGED
|
@@ -18,7 +18,9 @@ tags:
|
|
| 18 |
A "System 1" router: it reads a user prompt and returns its complexity score (0.01 to 1.00), its task type (19 classes)
|
| 19 |
and whether it needs tools. It never picks a model; the deployer maps these outputs to models with rules. The method,
|
| 20 |
the training data, the full results and the limitations are in the card of the main repository,
|
| 21 |
-
[scombat/s1d1-router](https://huggingface.co/scombat/s1d1-router).
|
|
|
|
|
|
|
| 22 |
|
| 23 |
- Variant: `v4_vocab32k-fp16`, full model with its vocabulary pruned to 32k tokens and a distillation pass.
|
| 24 |
- Format: PyTorch fp16 (safetensors), 308 MB.
|
|
|
|
| 18 |
A "System 1" router: it reads a user prompt and returns its complexity score (0.01 to 1.00), its task type (19 classes)
|
| 19 |
and whether it needs tools. It never picks a model; the deployer maps these outputs to models with rules. The method,
|
| 20 |
the training data, the full results and the limitations are in the card of the main repository,
|
| 21 |
+
[scombat/s1d1-router](https://huggingface.co/scombat/s1d1-router). An open, replayable bench of the variants against three local LLM
|
| 22 |
+
judges on 360 multilingual prompts, with confidence intervals and witnesses, is
|
| 23 |
+
[s1d1-bench](https://github.com/scombat/s1d1-bench).
|
| 24 |
|
| 25 |
- Variant: `v4_vocab32k-fp16`, full model with its vocabulary pruned to 32k tokens and a distillation pass.
|
| 26 |
- Format: PyTorch fp16 (safetensors), 308 MB.
|