Granite 8B Code Base By ibm-granite: Benchmarks, Features and Detailed Analysis. Insights on Granite 8B Code Base.

Arxiv:2405.04324 Autotrain compatible Code Dataset:bigcode/starcoderdata Dataset:codeparrot/github-code... Dataset:math-ai/stackmathqa Dataset:open-web-math/open-web... Granite Llama Model-index Region:us Safetensors Sharded Tensorflow

Model Card on HF 🤗: https://huggingface.co/ibm-granite/granite-8b-code-base

Granite 8B Code Base Benchmarks

LLME Score: 0.22029

^nn.n% — How the model compares to the reference models: Anthropic Sonnet 3.5 ("so35"), GPT-4o ("gpt4o") or GPT-4 ("gpt4").

What is the LLM Explorer Rank (Score)

Granite 8B Code Base (ibm-granite/granite-8b-code-base)

Granite 8B Code Base Parameters and Internals

Model Type

text generation

Use Cases

Applications:

Code generation, Code explanation, Code fixing, Unit tests, Documentation, Technical debt issues, Vulnerability detection, Code translation

Limitations:

Model not safety aligned, potential for problematic outputs

Training Details

Data Sources:

codeparrot/github-code-clean, bigcode/starcoderdata

Data Volume:

4 trillion tokens in phase 1, 500 billion tokens in phase 2

Methodology:

Two-phase training strategy with deduplication, HAP, PII, and malware filtering

Hardware Used:

NVIDIA A100 GPUs, NVIDIA H100 GPUs

Model Architecture:

Decoder-only model

Safety Evaluation

Ethical Considerations:

Use with ethical intentions and responsibly, model not safety-aligned, potential for problematic outputs.

LLM Name	Granite 8B Code Base
Repository 🤗	https://huggingface.co/ibm-granite/granite-8b-code-base
Model Size	8b
Required VRAM	16.2 GB
Updated	2024-09-01
Maintainer	ibm-granite
Model Type	llama
Model Files	5.0 GB: 1-of-4 4.9 GB: 2-of-4 5.0 GB: 3-of-4 1.3 GB: 4-of-4
Model Architecture	LlamaForCausalLM
License	apache-2.0
Context Length	4096
Model Max Length	4096
Transformers Version	4.41.0.dev0
Tokenizer Class	GPT2Tokenizer
Padding Token	<\|endoftext\|>
Vocabulary Size	49152
Torch Data Type	bfloat16

Quantized Models of the Granite 8B Code Base

Model	Likes	Downloads	VRAM
Granite 8B Code Instruct GGUF	6	12986	3 GB
Granite 8B Code Base 8bit	1	12	8 GB
Granite 8B Code Instruct 4bit	0	14	4 GB
Granite 8B Code Instruct 8bit	0	10	8 GB

Best Alternatives to Granite 8B Code Base

Best Alternatives	Context / RAM	Downloads	Likes
...a 3 8B Instruct Gradient 1048K	1024K / 16.1 GB	5133	681
R2	1024K / 16.1 GB	417	0
A6	1024K / 16.1 GB	152	0
D7	1024K / 16.1 GB	217	0
101	1024K / 16.1 GB	139	0
A2	1024K / 16.1 GB	68	0
A4	1024K / 16.1 GB	138	0
MrRoboto ProLong 8B V4i	1024K / 16.1 GB	66	1
D14	1024K / 16.1 GB	120	0
102	1024K / 16.1 GB	99	0

Note: green Score (e.g. "73.2") means that the model is better than ibm-granite/granite-8b-code-base.

Rank the Granite 8B Code Base Capabilities

🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟

Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation

What open-source LLMs or SLMs are you in search of? 44902 in total.

Email us: info@extractum.io. Our Privacy Policy | Terms and Conditions | Suggest an improvement.

Our Social Media →

Original data from HuggingFace, OpenCompass and various public git repos.

Release v20241227

Support LLM Explorer