> For the complete documentation index, see [llms.txt](https://docs.clore.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.clore.ai/guides/guides_v2-de/training/huggingface-transformers.md).

# HuggingFace Transformers

Verwende HuggingFace Transformers für NLP, Vision und Audio auf Clore.ai

Verwenden Sie die Transformers-Bibliothek für NLP, Vision und Audio auf der GPU.

{% hint style="success" %}
Alle Beispiele können auf GPU-Servern ausgeführt werden, die gemietet wurden über [CLORE.AI-Marktplatz](https://clore.ai/marketplace).
{% endhint %}

## Mieten auf CLORE.AI

1. Besuchen [CLORE.AI-Marktplatz](https://clore.ai/marketplace)
2. Filtern nach GPU-Typ, VRAM und Preis
3. Wähle **On-Demand** (Festpreis) oder **Spot** (Gebotspreis)
4. Konfiguriere deine Bestellung:
   * Docker-Image auswählen
   * Ports festlegen (TCP für SSH, HTTP für Web-UIs)
   * Bei Bedarf Umgebungsvariablen hinzufügen
   * Startbefehl eingeben
5. Zahlung auswählen: **CLORE**, **BTC**, oder **USDT/USDC**
6. Bestellung erstellen und auf die Bereitstellung warten

### Greife auf deinen Server zu

* Verbindungsdetails finden in **Meine Bestellungen**
* Web-Oberflächen: Verwende die HTTP-Port-URL
* SSH: `ssh -p <port> root@<proxy-address>`

## Was ist Transformers?

Hugging Face Transformers bietet:

* Über 100.000 vortrainierte Modelle
* Einfaches Laden von Modellen und Inferenz
* Unterstützung für Feinabstimmung
* Multimodale Funktionen

## Schnell bereitstellen

**Docker-Image:**

```
pytorch/pytorch:2.11.0-cuda12.8-cudnn9-devel
```

**Ports:**

```
22/tcp
```

**Befehl:**

```bash
pip install transformers accelerate datasets huggingface_hub
```

## Installation

```bash
pip install transformers[torch]
pip install accelerate  # Für große Modelle
pip install datasets    # Für Trainingsdaten
```

## Textgenerierung

### Grundlegende Generierung

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model_name = "mistralai/Mistral-7B-Instruct-v0.2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.float16,
    device_map="auto"
)

prompt = "Erkläre Quantencomputing in einfachen Worten:"
inputs = tokenizer(prompt, return_tensors="pt").to("cuda")

outputs = model.generate(
    **inputs,
    max_new_tokens=200,
    temperature=0.7,
    do_sample=True
)

response = tokenizer.decode(outputs[0], skip_special_tokens=True)
print(response)
```

### Chat-Modelle

```python
from transformers import pipeline

pipe = pipeline(
    "text-generation",
    model="meta-llama/Llama-2-7b-chat-hf",
    torch_dtype=torch.float16,
    device_map="auto"
)

messages = [
    {"role": "user", "content": "Was ist maschinelles Lernen?"}
]

outputs = pipe(
    messages,
    max_new_tokens=256,
    do_sample=True,
    temperature=0.7
)

print(outputs[0]["generated_text"][-1]["content"])

```

### Streaming

```python
from transformers import TextStreamer

streamer = TextStreamer(tokenizer)

model.generate(
    **inputs,
    max_new_tokens=200,
    streamer=streamer
)
```

## Quantisierung

### 4-Bit-Quantisierung

```python
from transformers import AutoModelForCausalLM, BitsAndBytesConfig
import torch

bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_quant_type="nf4",
    bnb_4bit_compute_dtype=torch.float16,
    bnb_4bit_use_double_quant=True
)

model = AutoModelForCausalLM.from_pretrained(
    "meta-llama/Llama-2-13b-hf",
    quantization_config=bnb_config,
    device_map="auto"
)
```

### 8-Bit-Quantisierung

```python
model = AutoModelForCausalLM.from_pretrained(
    "meta-llama/Llama-2-7b-hf",
    load_in_8bit=True,
    device_map="auto"
)
```

## Embeddings

```python
from transformers import AutoModel, AutoTokenizer
import torch

model_name = "sentence-transformers/all-MiniLM-L6-v2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModel.from_pretrained(model_name).cuda()

def get_embedding(text):
    inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True).to("cuda")
    with torch.no_grad():
        outputs = model(**inputs)
    # Mittelwert-Pooling
    embedding = outputs.last_hidden_state.mean(dim=1)
    return embedding

emb = get_embedding("Hello, world!")
print(f"Embedding shape: {emb.shape}")
```

## Bildklassifikation

```python
from transformers import pipeline
from PIL import Image

classifier = pipeline("image-classification", model="google/vit-base-patch16-224", device=0)

image = Image.open("cat.jpg")
results = classifier(image)

for result in results:
    print(f"{result['label']}: {result['score']:.4f}")
```

## Objekterkennung

```python
from transformers import pipeline
from PIL import Image

detector = pipeline("object-detection", model="facebook/detr-resnet-50", device=0)

image = Image.open("street.jpg")
results = detector(image)

for result in results:
    print(f"{result['label']}: {result['score']:.4f} at {result['box']}")
```

## Bildsegmentierung

```python
from transformers import pipeline
from PIL import Image

segmenter = pipeline("image-segmentation", model="facebook/maskformer-swin-base-ade", device=0)

image = Image.open("scene.jpg")
results = segmenter(image)

for segment in results:
    print(f"{segment['label']}: score {segment['score']:.4f}")
```

## Spracherkennung

```python
from transformers import pipeline

transcriber = pipeline(
    "automatic-speech-recognition",
    model="openai/whisper-large-v3",
    device=0
)

result = transcriber("audio.mp3")
print(result["text"])

```

## Text-zu-Sprache

```python
from transformers import pipeline
import scipy

synthesizer = pipeline("text-to-speech", model="microsoft/speecht5_tts", device=0)

speech = synthesizer("Hallo, dies ist ein Test von Text-zu-Sprache.")

scipy.io.wavfile.write("output.wav", rate=speech["sampling_rate"], data=speech["audio"])

```

## Feinabstimmung

### Datensatz vorbereiten

```python
from datasets import load_dataset

dataset = load_dataset("imdb")
train_dataset = dataset["train"].select(range(1000))
eval_dataset = dataset["test"].select(range(200))
```

### Training

```python
from transformers import (
    AutoModelForSequenceClassification,
    AutoTokenizer,
    TrainingArguments,
    Trainer
)

model_name = "bert-base-uncased"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name, num_labels=2)

def tokenize_function(examples):
    return tokenizer(examples["text"], padding="max_length", truncation=True)

tokenized_train = train_dataset.map(tokenize_function, batched=True)
tokenized_eval = eval_dataset.map(tokenize_function, batched=True)

training_args = TrainingArguments(
    output_dir="./results",
    evaluation_strategy="epoch",
    learning_rate=2e-5,
    per_device_train_batch_size=16,
    per_device_eval_batch_size=16,
    num_train_epochs=3,
    weight_decay=0.01,
    fp16=True,
)

trainer = Trainer(
    model=model,
    args=training_args,
    train_dataset=tokenized_train,
    eval_dataset=tokenized_eval,
)

trainer.train()
```

## Pipeline-Aufgaben

| Aufgabe            | Pipeline-Name                  |
| ------------------ | ------------------------------ |
| Textgenerierung    | `text-generation`              |
| Maskenfüllung      | `fill-mask`                    |
| Zusammenfassung    | `summarization`                |
| Übersetzung        | `translation`                  |
| Fragebeantwortung  | `question-answering`           |
| Sentimentanalyse   | `sentiment-analysis`           |
| Bildklassifikation | `image-classification`         |
| Objekterkennung    | `object-detection`             |
| Spracherkennung    | `automatic-speech-recognition` |
| Text-zu-Sprache    | `text-to-speech`               |

## Multi-GPU

```python
from transformers import AutoModelForCausalLM

# Automatische Geräteplatzierung
model = AutoModelForCausalLM.from_pretrained(
    "meta-llama/Llama-2-70b-hf",
    device_map="auto",
    torch_dtype=torch.float16
)

# Manuelle Gerätezuordnung
device_map = {
    "model.embed_tokens": 0,
    "model.layers.0": 0,
    "model.layers.1": 0,
    # ...
    "model.layers.39": 1,
    "model.norm": 1,
    "lm_head": 1
}

model = AutoModelForCausalLM.from_pretrained(
    "model_name",
    device_map=device_map
)
```

## Speicheroptimierung

```python

# Flash Attention 2
model = AutoModelForCausalLM.from_pretrained(
    "meta-llama/Llama-2-7b-hf",
    torch_dtype=torch.float16,
    attn_implementation="flash_attention_2",
    device_map="auto"
)

# Gradient-Checkpointing
model.gradient_checkpointing_enable()
```

## Modell-Hub

### Modelle herunterladen

```python
from huggingface_hub import snapshot_download

snapshot_download(
    repo_id="meta-llama/Llama-2-7b-hf",
    local_dir="./llama-2-7b"
)
```

### Modelle hochladen

```python
from huggingface_hub import HfApi

api = HfApi()
api.upload_folder(
    folder_path="./my_model",
    repo_id="username/my-model",
    repo_type="model"
)
```

## Leistungstipps

| Tipp                                 | Effekt                  |
| ------------------------------------ | ----------------------- |
| Verwende `torch_dtype=torch.float16` | 50 % weniger Speicher   |
| Flash Attention 2 aktivieren         | 2× schnellere Attention |
| Quantisierung verwenden              | 75 % weniger Speicher   |
| Batch-Inferenz                       | Höherer Durchsatz       |

## Fehlerbehebung

## Kostenschätzung

Übliche CLORE.AI-Marktplatzpreise (Stand 2024):

| GPU       | Stundensatz | Tagessatz | 4-Stunden-Sitzung |
| --------- | ----------- | --------- | ----------------- |
| RTX 3060  | \~$0.03     | \~$0.70   | \~$0.12           |
| RTX 3090  | \~$0.06     | \~$1.50   | \~$0.25           |
| RTX 4090  | \~$0.10     | \~$2.30   | \~$0.40           |
| A100 40GB | \~$0.17     | \~$4.00   | \~$0.70           |
| A100 80GB | \~$0.25     | \~$6.00   | \~$1.00           |

*Die Preise variieren je nach Anbieter und Nachfrage. Prüfen Sie* [*CLORE.AI-Marktplatz*](https://clore.ai/marketplace) *für aktuelle Preise.*

**Geld sparen:**

* Nutzen Sie den **Spot** Markt für unterbrechbare Arbeit — etwa ein Drittel der Server bietet Spot-Preise unter dem On-Demand-Preis (Median ca. 13 % Rabatt), der Rest ist gleichauf
* Bezahlen Sie mit **CLORE** Tokens
* Vergleichen Sie Preise zwischen verschiedenen Anbietern

## Nächste Schritte

* vLLM Inference - Produktionsbereitstellung
* [LLMs feinabstimmen](/guides/guides_v2-de/training/finetune-llm.md) - LoRA-Training
* [DeepSpeed-Training](/guides/guides_v2-de/training/deepspeed-training.md) - Verteiltes Training


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.clore.ai/guides/guides_v2-de/training/huggingface-transformers.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.