Instructions to use QuantFactory/granite-3b-code-base-128k-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use QuantFactory/granite-3b-code-base-128k-GGUF with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="QuantFactory/granite-3b-code-base-128k-GGUF")

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("QuantFactory/granite-3b-code-base-128k-GGUF", dtype="auto")

llama-cpp-python

How to use QuantFactory/granite-3b-code-base-128k-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="QuantFactory/granite-3b-code-base-128k-GGUF",
	filename="granite-3b-code-base-128k.Q2_K.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use QuantFactory/granite-3b-code-base-128k-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M

Use Docker

docker model run hf.co/QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use QuantFactory/granite-3b-code-base-128k-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "QuantFactory/granite-3b-code-base-128k-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "QuantFactory/granite-3b-code-base-128k-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M

SGLang

How to use QuantFactory/granite-3b-code-base-128k-GGUF with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "QuantFactory/granite-3b-code-base-128k-GGUF" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "QuantFactory/granite-3b-code-base-128k-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "QuantFactory/granite-3b-code-base-128k-GGUF" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "QuantFactory/granite-3b-code-base-128k-GGUF",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Ollama
How to use QuantFactory/granite-3b-code-base-128k-GGUF with Ollama:
```
ollama run hf.co/QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M
```

Unsloth Studio new

How to use QuantFactory/granite-3b-code-base-128k-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for QuantFactory/granite-3b-code-base-128k-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for QuantFactory/granite-3b-code-base-128k-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for QuantFactory/granite-3b-code-base-128k-GGUF to start chatting

Docker Model Runner
How to use QuantFactory/granite-3b-code-base-128k-GGUF with Docker Model Runner:
```
docker model run hf.co/QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M
```

Lemonade

How to use QuantFactory/granite-3b-code-base-128k-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull QuantFactory/granite-3b-code-base-128k-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.granite-3b-code-base-128k-GGUF-Q4_K_M

List all available models

lemonade list

munish0838 commited on Sep 7, 2024

Commit

9ca050f

verified ·

1 Parent(s): 0fb69a3

Upload README.md with huggingface_hub

Browse files

Files changed (1) hide show

README.md +261 -0

README.md ADDED Viewed

	@@ -0,0 +1,261 @@

+---
+pipeline_tag: text-generation
+inference: false
+license: apache-2.0
+datasets:
+- codeparrot/github-code-clean
+- bigcode/starcoderdata
+# - Stackexchange
+# - CommonCrawl
+- open-web-math/open-web-math
+- math-ai/StackMathQA
+# - Arxiv
+# - Wikipedia
+# - conceptofmind/FLAN_2022 # Original link is broken, we used IBM's filtered version
+metrics:
+- code_eval
+library_name: transformers
+tags:
+- code
+- granite
+model-index:
+- name: granite-3b-code-base-128k
+  results:
+  - task:
+      type: text-generation
+    dataset:
+        type: bigcode/humanevalpack
+        name: HumanEvalSynthesis (Python)
+    metrics:
+    - name: pass@1
+      type: pass@1
+      value: 36.0
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: bigcode/humanevalpack
+        name: HumanEvalSynthesis (Average)
+    metrics:
+    - name: pass@1
+      type: pass@1
+      value: 30.5
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: bigcode/humanevalpack
+        name:  HumanEvalExplain (Average)
+    metrics:
+    - name: pass@1
+      type: pass@1
+      value: 22.4
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: bigcode/humanevalpack
+        name:  HumanEvalFix (Average)
+    metrics:
+    - name: pass@1
+      type: pass@1
+      value: 19.9
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: repoqa
+        name:  RepoQA (Python@16K)
+    metrics:
+    - name: pass@1 (thresh=0.5)
+      type: pass@1 (thresh=0.5)
+      value: 40.0
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: repoqa
+        name:  RepoQA (C++@16K)
+    metrics:
+    - name: pass@1 (thresh=0.5)
+      type: pass@1 (thresh=0.5)
+      value: 36.0
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: repoqa
+        name:  RepoQA (Java@16K)
+    metrics:
+    - name: pass@1 (thresh=0.5)
+      type: pass@1 (thresh=0.5)
+      value: 37.0
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: repoqa
+        name:  RepoQA (TypeScript@16K)
+    metrics:
+    - name: pass@1 (thresh=0.5)
+      type: pass@1 (thresh=0.5)
+      value: 27.0
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: repoqa
+        name:  RepoQA (Rust@16K)
+    metrics:
+    - name: pass@1 (thresh=0.5)
+      type: pass@1 (thresh=0.5)
+      value: 29.0
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: lcc
+        name:  LCC (Balanced)
+    metrics:
+    - name: Exact Match@4K
+      type: Exact Match@4K
+      value: 54.6
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: lcc
+        name:  LCC (Balanced)
+    metrics:
+    - name: Exact Match@8K
+      type: Exact Match@8K
+      value: 56.8
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: lcc
+        name:  LCC (Balanced)
+    metrics:
+    - name: Exact Match@16K
+      type: Exact Match@16K
+      value: 52.2
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: lcc
+        name:  LCC (Balanced)
+    metrics:
+    - name: Exact Match@32K
+      type: Exact Match@32K
+      value: 57.8
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: repobench
+        name: RepoBench-P (Balanced)
+    metrics:
+    - name: Exact Match@4K
+      type: Exact Match@4K
+      value: 39.8
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: repobench
+        name: RepoBench-P (Balanced)
+    metrics:
+    - name: Exact Match@8K
+      type: Exact Match@8K
+      value: 46.8
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: repobench
+        name: RepoBench-P (Balanced)
+    metrics:
+    - name: Exact Match@16K
+      type: Exact Match@16K
+      value: 43.1
+      verified: false
+  - task:
+      type: text-generation
+    dataset:
+        type: repobench
+        name: RepoBench-Pn(Balanced)
+    metrics:
+    - name: Exact Match@32K
+      type: Exact Match@32K
+      value: 45.3
+      verified: false
+---
+![](https://lh7-rt.googleusercontent.com/docsz/AD_4nXeiuCm7c8lEwEJuRey9kiVZsRn2W-b4pWlu3-X534V3YmVuVc2ZL-NXg2RkzSOOS2JXGHutDuyyNAUtdJI65jGTo8jT9Y99tMi4H4MqL44Uc5QKG77B0d6-JfIkZHFaUA71-RtjyYZWVIhqsNZcx8-OMaA?key=xt3VSDoCbmTY7o-cwwOFwQ)
+# QuantFactory/granite-3b-code-base-128k-GGUF
+This is quantized version of [ibm-granite/granite-3b-code-base-128k](https://huggingface.co/ibm-granite/granite-3b-code-base-128k) created using llama.cpp
+# Original Model Card
+![image/png](https://cdn-uploads.huggingface.co/production/uploads/62cd5057674cdb524450093d/1hzxoPwqkBJXshKVVe6_9.png)
+# Granite-3B-Code-Base-128K
+## Model Summary
+**Granite-3B-Code-Base-128K** extends the context length of Granite-3B-Code-Base from 2K to 128K with continual pretraining using the original training data but with repository-level file packing and per-language length upsampling, that we found to be critical for long-context pretraining.
+We adopt an progressive training strategy where we doubled the context window until it reached the desired length of 128K by appropriately adjusting RoPE theta. We trained on 4B tokens total for all stages, which is only 0.1% of Granite-3B-Code-Base's original pre-training data.
+- **Developers:** IBM Research
+- **GitHub Repository:** [ibm-granite/granite-code-models](https://github.com/ibm-granite/granite-code-models)
+- **Paper:** [Scaling Granite Code Models to 128K Context](https://arxiv.org/abs/2405.04324)
+- **Release Date**: July 18th, 2024
+- **License:** [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0).
+## Usage
+### Intended use
+Prominent enterprise use cases of LLMs in software engineering productivity with 128K context length support that includes code generation, code explanation, code fixing, generating unit tests, generating documentation, addressing technical debt issues, vulnerability detection, code translation, and more. All Granite Code Base models, including the **3B parameter model**, are able to handle these tasks as they were trained on a large amount of code data from 116 programming languages.
+### Generation
+This is a simple example of how to use **Granite-3B-Code-Base-128K** model.
+```python
+import torch
+from transformers import AutoModelForCausalLM, AutoTokenizer
+device = "cuda" # or "cpu"
+model_path = "ibm-granite/granite-3b-code-base-128k"
+tokenizer = AutoTokenizer.from_pretrained(model_path)
+# drop device_map if running on CPU
+model = AutoModelForCausalLM.from_pretrained(model_path, device_map=device)
+model.eval()
+# change input text as desired
+input_text = "def generate():"
+# tokenize the text
+input_tokens = tokenizer(input_text, return_tensors="pt")
+# transfer tokenized inputs to the device
+for i in input_tokens:
+    input_tokens[i] = input_tokens[i].to(device)
+# generate output tokens
+output = model.generate(**input_tokens)
+# decode output tokens into text
+output = tokenizer.batch_decode(output)
+# loop over the batch to print, in this example the batch size is 1
+for i in output:
+    print(i)
+```
+## Training Data
+Starting from the base Granite model, this model was further pretrained on repository-level code data with per-language context-length oversampling, allowing it to effectively utilize up to 128K tokens of context. This continued training stage focused on a curated selection of programming languages, such as Python, C, C++, Go, Java, JavaScript, and TypeScript.
+## Infrastructure
+We train the Granite Code models using two of IBM's super computing clusters, namely Vela and Blue Vela, both outfitted with NVIDIA A100 and H100 GPUs respectively. These clusters provide a scalable and efficient infrastructure for training our models over thousands of GPUs.
+## Ethical Considerations and Limitations
+The use of Large Language Models involves risks and ethical considerations people must be aware of. Regarding code generation, caution is urged against complete reliance on specific code models for crucial decisions or impactful information as the generated code is not guaranteed to work as intended. **Granite-3B-Code-Base-128K** model is not the exception in this regard. Even though this model is suited for multiple code-related tasks, it has not undergone any safety alignment, there it may produce problematic outputs. Additionally, it remains uncertain whether smaller models might exhibit increased susceptibility to hallucination in generation scenarios by copying source code verbatim from the training dataset due to their reduced sizes and memorization capacities. This aspect is currently an active area of research, and we anticipate more rigorous exploration, comprehension, and mitigations in this domain. Regarding ethics, a latent risk associated with all Large Language Models is their malicious utilization. We urge the community to use **Granite-3B-Code-Base-128K** model with ethical intentions and in a responsible way.