metadata
pipeline_tag: text-generation
inference: true
license: apache-2.0
datasets:
- codeparrot/github-code-clean
- bigcode/starcoderdata
- open-web-math/open-web-math
- math-ai/StackMathQA
metrics:
- code_eval
library_name: transformers
tags:
- code
- granite
- TensorBlock
- GGUF
base_model: ibm-granite/granite-34b-code-base-8k
model-index:
- name: granite-34b-code-base-8k
results:
- task:
type: text-generation
dataset:
name: MBPP
type: mbpp
metrics:
- type: pass@1
value: 47.2
name: pass@1
- task:
type: text-generation
dataset:
name: MBPP+
type: evalplus/mbppplus
metrics:
- type: pass@1
value: 53.1
name: pass@1
- task:
type: text-generation
dataset:
name: HumanEvalSynthesis(Python)
type: bigcode/humanevalpack
metrics:
- type: pass@1
value: 48.2
name: pass@1
- type: pass@1
value: 54.9
name: pass@1
- type: pass@1
value: 61.6
name: pass@1
- type: pass@1
value: 40.2
name: pass@1
- type: pass@1
value: 50
name: pass@1
- type: pass@1
value: 39.6
name: pass@1
- type: pass@1
value: 42.7
name: pass@1
- type: pass@1
value: 26.2
name: pass@1
- type: pass@1
value: 47
name: pass@1
- type: pass@1
value: 26.8
name: pass@1
- type: pass@1
value: 36.6
name: pass@1
- type: pass@1
value: 25
name: pass@1
- type: pass@1
value: 20.1
name: pass@1
- type: pass@1
value: 30.5
name: pass@1
- type: pass@1
value: 40.9
name: pass@1
- type: pass@1
value: 34.1
name: pass@1
- type: pass@1
value: 39
name: pass@1
- type: pass@1
value: 12.2
name: pass@1
Feedback and support: TensorBlock's Twitter/X, Telegram Group and Discord server
ibm-granite/granite-34b-code-base-8k - GGUF
This repo contains GGUF format model files for ibm-granite/granite-34b-code-base-8k.
The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b4011.
Our projects
| Awesome MCP Servers | TensorBlock Studio |
|---|---|
 |
 |
| A comprehensive collection of Model Context Protocol (MCP) servers. | A lightweight, open, and extensible multi-LLM interaction studio. |
| 👀 See what we built 👀 | 👀 See what we built 👀 |
Model file specification
| Filename | Quant type | File Size | Description |
|---|---|---|---|
| granite-34b-code-base-8k-Q2_K.gguf | Q2_K | 12.207 GB | smallest, significant quality loss - not recommended for most purposes |
| granite-34b-code-base-8k-Q3_K_S.gguf | Q3_K_S | 13.791 GB | very small, high quality loss |
| granite-34b-code-base-8k-Q3_K_M.gguf | Q3_K_M | 16.361 GB | very small, high quality loss |
| granite-34b-code-base-8k-Q3_K_L.gguf | Q3_K_L | 18.207 GB | small, substantial quality loss |
| granite-34b-code-base-8k-Q4_0.gguf | Q4_0 | 17.917 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| granite-34b-code-base-8k-Q4_K_S.gguf | Q4_K_S | 18.110 GB | small, greater quality loss |
| granite-34b-code-base-8k-Q4_K_M.gguf | Q4_K_M | 19.915 GB | medium, balanced quality - recommended |
| granite-34b-code-base-8k-Q5_0.gguf | Q5_0 | 21.800 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| granite-34b-code-base-8k-Q5_K_S.gguf | Q5_K_S | 21.800 GB | large, low quality loss - recommended |
| granite-34b-code-base-8k-Q5_K_M.gguf | Q5_K_M | 23.050 GB | large, very low quality loss - recommended |
| granite-34b-code-base-8k-Q6_K.gguf | Q6_K | 25.926 GB | very large, extremely low quality loss |
| granite-34b-code-base-8k-Q8_0.gguf | Q8_0 | 33.518 GB | very large, extremely low quality loss - not recommended |
Downloading instruction
Command line
Firstly, install Huggingface Client
pip install -U "huggingface_hub[cli]"
Then, downoad the individual model file the a local directory
huggingface-cli download tensorblock/granite-34b-code-base-8k-GGUF --include "granite-34b-code-base-8k-Q2_K.gguf" --local-dir MY_LOCAL_DIR
If you wanna download multiple model files with a pattern (e.g., *Q4_K*gguf), you can try:
huggingface-cli download tensorblock/granite-34b-code-base-8k-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'