Granite Code Models
Collection
A series of code models trained by IBM licensed under Apache 2.0 license. We release both the base pretrained and instruct models.
•
23 items
•
Updated
•
178
ibm-granite/granite-34b-code-instruct-8k-GGUF
This is the Q4_K_M converted version of the original ibm-granite/granite-34b-code-instruct-8k.
Refer to the original model card for more details.
Use with llama.cpp
git clone https://github.com/ggerganov/llama.cpp
cd llama.cpp
# install
make
# run generation
./main -m granite-34b-code-instruct-8k-GGUF/granite-34b-code-instruct.Q4_K_M.gguf -n 128 -p "def generate_random(x: int):" --color
- Downloads last month
- 131
Model tree for ibm-granite/granite-34b-code-instruct-8k-GGUF
Base model
ibm-granite/granite-34b-code-base-8kDatasets used to train ibm-granite/granite-34b-code-instruct-8k-GGUF
Collection including ibm-granite/granite-34b-code-instruct-8k-GGUF
Evaluation results
- pass@1 on HumanEvalSynthesis(Python)self-reported62.200
- pass@1 on HumanEvalSynthesis(JavaScript)self-reported56.700
- pass@1 on HumanEvalSynthesis(Java)self-reported62.800
- pass@1 on HumanEvalSynthesis(Go)self-reported47.600
- pass@1 on HumanEvalSynthesis(C++)self-reported57.900
- pass@1 on HumanEvalSynthesis(Rust)self-reported41.500
- pass@1 on HumanEvalExplain(Python)self-reported53.000
- pass@1 on HumanEvalExplain(JavaScript)self-reported45.100
- pass@1 on HumanEvalExplain(Java)self-reported50.600
- pass@1 on HumanEvalExplain(Go)self-reported36.000