--- base_model: OpenLLM-Ro/RoLlama3.1-8b-Instruct-2025-04-23 datasets: - OpenLLM-Ro/ro_sft_alpaca - OpenLLM-Ro/ro_sft_alpaca_gpt4 - OpenLLM-Ro/ro_sft_dolly - OpenLLM-Ro/ro_sft_selfinstruct_gpt4 - OpenLLM-Ro/ro_sft_norobots - OpenLLM-Ro/ro_sft_orca - OpenLLM-Ro/ro_sft_camel - OpenLLM-Ro/ro_sft_oasst - OpenLLM-Ro/ro_sft_ultrachat - OpenLLM-Ro/ro_sft_magpie_mt - OpenLLM-Ro/ro_sft_magpie_reasoning language: - ro license: cc-by-nc-4.0 tags: - llama-cpp - gguf-my-repo model-index: - name: OpenLLM-Ro/RoLlama3.1-8b-Instruct-2025-04-23 results: - task: type: text-generation dataset: name: RoMT-Bench type: RoMT-Bench metrics: - type: Score value: 6.43 name: Score - type: Score value: 6.78 name: First turn - type: Score value: 6.09 name: Second turn - task: type: text-generation dataset: name: RoCulturaBench type: RoCulturaBench metrics: - type: Score value: 4.28 name: Score - task: type: text-generation dataset: name: Romanian_Academic_Benchmarks type: Romanian_Academic_Benchmarks metrics: - type: accuracy value: 53.36 name: Average accuracy - task: type: text-generation dataset: name: OpenLLM-Ro/ro_arc_challenge type: OpenLLM-Ro/ro_arc_challenge metrics: - type: accuracy value: 48.97 name: Average accuracy - type: accuracy value: 45.24 name: 0-shot - type: accuracy value: 47.67 name: 1-shot - type: accuracy value: 49.36 name: 3-shot - type: accuracy value: 50.13 name: 5-shot - type: accuracy value: 50.81 name: 10-shot - type: accuracy value: 50.64 name: 25-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_mmlu type: OpenLLM-Ro/ro_mmlu metrics: - type: accuracy value: 55.17 name: Average accuracy - type: accuracy value: 54.23 name: 0-shot - type: accuracy value: 56.36 name: 1-shot - type: accuracy value: 55.34 name: 3-shot - type: accuracy value: 54.74 name: 5-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_winogrande type: OpenLLM-Ro/ro_winogrande metrics: - type: accuracy value: 66.52 name: Average accuracy - type: accuracy value: 64.96 name: 0-shot - type: accuracy value: 66.77 name: 1-shot - type: accuracy value: 67.09 name: 3-shot - type: accuracy value: 67.25 name: 5-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_hellaswag type: OpenLLM-Ro/ro_hellaswag metrics: - type: accuracy value: 60.73 name: Average accuracy - type: accuracy value: 59.72 name: 0-shot - type: accuracy value: 60.3 name: 1-shot - type: accuracy value: 60.87 name: 3-shot - type: accuracy value: 61.14 name: 5-shot - type: accuracy value: 61.63 name: 10-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_gsm8k type: OpenLLM-Ro/ro_gsm8k metrics: - type: accuracy value: 42.03 name: Average accuracy - type: accuracy value: 30.86 name: 1-shot - type: accuracy value: 43.9 name: 3-shot - type: accuracy value: 51.33 name: 5-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_truthfulqa type: OpenLLM-Ro/ro_truthfulqa metrics: - type: accuracy value: 46.71 name: Average accuracy - task: type: text-generation dataset: name: LaRoSeDa_binary type: LaRoSeDa_binary metrics: - type: macro-f1 value: 95.32 name: Average macro-f1 - type: macro-f1 value: 90.97 name: 0-shot - type: macro-f1 value: 95.53 name: 1-shot - type: macro-f1 value: 97.1 name: 3-shot - type: macro-f1 value: 97.67 name: 5-shot - task: type: text-generation dataset: name: LaRoSeDa_multiclass type: LaRoSeDa_multiclass metrics: - type: macro-f1 value: 60.84 name: Average macro-f1 - type: macro-f1 value: 63.2 name: 0-shot - type: macro-f1 value: 64.47 name: 1-shot - type: macro-f1 value: 55.88 name: 3-shot - type: macro-f1 value: 59.8 name: 5-shot - task: type: text-generation dataset: name: WMT_EN-RO type: WMT_EN-RO metrics: - type: bleu value: 23.18 name: Average bleu - type: bleu value: 4.92 name: 0-shot - type: bleu value: 28.01 name: 1-shot - type: bleu value: 30.16 name: 3-shot - type: bleu value: 29.61 name: 5-shot - task: type: text-generation dataset: name: WMT_RO-EN type: WMT_RO-EN metrics: - type: bleu value: 25.11 name: Average bleu - type: bleu value: 1.43 name: 0-shot - type: bleu value: 24.78 name: 1-shot - type: bleu value: 37.31 name: 3-shot - type: bleu value: 36.93 name: 5-shot - task: type: text-generation dataset: name: XQuAD type: XQuAD metrics: - type: exact_match value: 10.74 name: Average exact_match - type: f1 value: 19.75 name: Average f1 - task: type: text-generation dataset: name: STS type: STS metrics: - type: spearman value: 73.53 name: Average spearman - type: pearson value: 74.93 name: Average pearson - task: type: text-generation dataset: name: XQuAD_EM type: XQuAD_EM metrics: - type: exact_match value: 11.18 name: 0-shot - type: exact_match value: 26.47 name: 1-shot - type: exact_match value: 3.95 name: 3-shot - type: exact_match value: 1.34 name: 5-shot - task: type: text-generation dataset: name: XQuAD_F1 type: XQuAD_F1 metrics: - type: f1 value: 25.76 name: 0-shot - type: f1 value: 39.25 name: 1-shot - type: f1 value: 8.4 name: 3-shot - type: f1 value: 5.58 name: 5-shot - task: type: text-generation dataset: name: STS_Spearman type: STS_Spearman metrics: - type: spearman value: 73.52 name: 1-shot - type: spearman value: 74.02 name: 3-shot - type: spearman value: 73.06 name: 5-shot - task: type: text-generation dataset: name: STS_Pearson type: STS_Pearson metrics: - type: pearson value: 75.81 name: 1-shot - type: pearson value: 74.54 name: 3-shot - type: pearson value: 74.43 name: 5-shot --- # LinigDrake2875/RoLlama3.1-8b-Instruct-2025-04-23-Q4_K_M-GGUF This model was converted to GGUF format from [`OpenLLM-Ro/RoLlama3.1-8b-Instruct-2025-04-23`](https://huggingface.co/OpenLLM-Ro/RoLlama3.1-8b-Instruct-2025-04-23) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space. Refer to the [original model card](https://huggingface.co/OpenLLM-Ro/RoLlama3.1-8b-Instruct-2025-04-23) for more details on the model. ## Use with llama.cpp Install llama.cpp through brew (works on Mac and Linux) ```bash brew install llama.cpp ``` Invoke the llama.cpp server or the CLI. ### CLI: ```bash llama-cli --hf-repo LinigDrake2875/RoLlama3.1-8b-Instruct-2025-04-23-Q4_K_M-GGUF --hf-file rollama3.1-8b-instruct-2025-04-23-q4_k_m.gguf -p "The meaning to life and the universe is" ``` ### Server: ```bash llama-server --hf-repo LinigDrake2875/RoLlama3.1-8b-Instruct-2025-04-23-Q4_K_M-GGUF --hf-file rollama3.1-8b-instruct-2025-04-23-q4_k_m.gguf -c 2048 ``` Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well. Step 1: Clone llama.cpp from GitHub. ``` git clone https://github.com/ggerganov/llama.cpp ``` Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux). ``` cd llama.cpp && LLAMA_CURL=1 make ``` Step 3: Run inference through the main binary. ``` ./llama-cli --hf-repo LinigDrake2875/RoLlama3.1-8b-Instruct-2025-04-23-Q4_K_M-GGUF --hf-file rollama3.1-8b-instruct-2025-04-23-q4_k_m.gguf -p "The meaning to life and the universe is" ``` or ``` ./llama-server --hf-repo LinigDrake2875/RoLlama3.1-8b-Instruct-2025-04-23-Q4_K_M-GGUF --hf-file rollama3.1-8b-instruct-2025-04-23-q4_k_m.gguf -c 2048 ```