metadata
pipeline_tag: text-generation
inference: false
license: apache-2.0
library_name: transformers
tags:
- language
- granite-3.0
- TensorBlock
- GGUF
base_model: ibm-granite/granite-3.0-1b-a400m-base
model-index:
- name: granite-3.0-1b-a400m-base
results:
- task:
type: text-generation
dataset:
name: MMLU
type: human-exams
metrics:
- type: pass@1
value: 25.69
name: pass@1
- type: pass@1
value: 11.38
name: pass@1
- type: pass@1
value: 19.96
name: pass@1
- task:
type: text-generation
dataset:
name: WinoGrande
type: commonsense
metrics:
- type: pass@1
value: 62.43
name: pass@1
- type: pass@1
value: 39
name: pass@1
- type: pass@1
value: 35.76
name: pass@1
- type: pass@1
value: 75.35
name: pass@1
- type: pass@1
value: 64.92
name: pass@1
- type: pass@1
value: 39.49
name: pass@1
- task:
type: text-generation
dataset:
name: BoolQ
type: reading-comprehension
metrics:
- type: pass@1
value: 65.44
name: pass@1
- type: pass@1
value: 17.78
name: pass@1
- task:
type: text-generation
dataset:
name: ARC-C
type: reasoning
metrics:
- type: pass@1
value: 38.14
name: pass@1
- type: pass@1
value: 24.41
name: pass@1
- type: pass@1
value: 29.84
name: pass@1
- type: pass@1
value: 33.99
name: pass@1
- task:
type: text-generation
dataset:
name: HumanEval
type: code
metrics:
- type: pass@1
value: 21.95
name: pass@1
- type: pass@1
value: 23.2
name: pass@1
- task:
type: text-generation
dataset:
name: GSM8K
type: math
metrics:
- type: pass@1
value: 19.26
name: pass@1
- type: pass@1
value: 8.96
name: pass@1
Feedback and support: TensorBlock's Twitter/X, Telegram Group and Discord server
ibm-granite/granite-3.0-1b-a400m-base - GGUF
This repo contains GGUF format model files for ibm-granite/granite-3.0-1b-a400m-base.
The files were quantized using machines provided by TensorBlock, and they are compatible with llama.cpp as of commit b4011.
Prompt template
Model file specification
| Filename | Quant type | File Size | Description |
|---|---|---|---|
| granite-3.0-1b-a400m-base-Q2_K.gguf | Q2_K | 0.529 GB | smallest, significant quality loss - not recommended for most purposes |
| granite-3.0-1b-a400m-base-Q3_K_S.gguf | Q3_K_S | 0.619 GB | very small, high quality loss |
| granite-3.0-1b-a400m-base-Q3_K_M.gguf | Q3_K_M | 0.680 GB | very small, high quality loss |
| granite-3.0-1b-a400m-base-Q3_K_L.gguf | Q3_K_L | 0.733 GB | small, substantial quality loss |
| granite-3.0-1b-a400m-base-Q4_0.gguf | Q4_0 | 0.797 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
| granite-3.0-1b-a400m-base-Q4_K_S.gguf | Q4_K_S | 0.803 GB | small, greater quality loss |
| granite-3.0-1b-a400m-base-Q4_K_M.gguf | Q4_K_M | 0.850 GB | medium, balanced quality - recommended |
| granite-3.0-1b-a400m-base-Q5_0.gguf | Q5_0 | 0.963 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
| granite-3.0-1b-a400m-base-Q5_K_S.gguf | Q5_K_S | 0.963 GB | large, low quality loss - recommended |
| granite-3.0-1b-a400m-base-Q5_K_M.gguf | Q5_K_M | 0.991 GB | large, very low quality loss - recommended |
| granite-3.0-1b-a400m-base-Q6_K.gguf | Q6_K | 1.140 GB | very large, extremely low quality loss |
| granite-3.0-1b-a400m-base-Q8_0.gguf | Q8_0 | 1.476 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-3.0-1b-a400m-base-GGUF --include "granite-3.0-1b-a400m-base-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-3.0-1b-a400m-base-GGUF --local-dir MY_LOCAL_DIR --local-dir-use-symlinks False --include='*Q4_K*gguf'