--- base_model: OpenLLM-Ro/RoLlama3-8b-Instruct-DPO datasets: - OpenLLM-Ro/ro_dpo_helpsteer language: - ro license: cc-by-nc-4.0 tags: - llama-cpp - gguf-my-repo model-index: - name: OpenLLM-Ro/RoLlama3-8b-Instruct-DPO-2024-10-09 results: - task: type: text-generation dataset: name: RoMT-Bench type: RoMT-Bench metrics: - type: Score value: 5.87 name: Score - type: Score value: 6.22 name: First turn - type: Score value: 5.49 name: Second turn - task: type: text-generation dataset: name: RoCulturaBench type: RoCulturaBench metrics: - type: Score value: 4.4 name: Score - task: type: text-generation dataset: name: Romanian_Academic_Benchmarks type: Romanian_Academic_Benchmarks metrics: - type: accuracy value: 49.96 name: Average accuracy - task: type: text-generation dataset: name: OpenLLM-Ro/ro_arc_challenge type: OpenLLM-Ro/ro_arc_challenge metrics: - type: accuracy value: 46.29 name: Average accuracy - type: accuracy value: 44.56 name: 0-shot - type: accuracy value: 45.42 name: 1-shot - type: accuracy value: 46.1 name: 3-shot - type: accuracy value: 46.27 name: 5-shot - type: accuracy value: 46.96 name: 10-shot - type: accuracy value: 48.41 name: 25-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_mmlu type: OpenLLM-Ro/ro_mmlu metrics: - type: accuracy value: 53.29 name: Average accuracy - type: accuracy value: 52.33 name: 0-shot - type: accuracy value: 52.86 name: 1-shot - type: accuracy value: 54.06 name: 3-shot - type: accuracy value: 53.9 name: 5-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_winogrande type: OpenLLM-Ro/ro_winogrande metrics: - type: accuracy value: 65.57 name: Average accuracy - type: accuracy value: 64.33 name: 0-shot - type: accuracy value: 64.72 name: 1-shot - type: accuracy value: 66.3 name: 3-shot - type: accuracy value: 66.93 name: 5-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_hellaswag type: OpenLLM-Ro/ro_hellaswag metrics: - type: accuracy value: 58.15 name: Average accuracy - type: accuracy value: 57.45 name: 0-shot - type: accuracy value: 57.65 name: 1-shot - type: accuracy value: 58.18 name: 3-shot - type: accuracy value: 58.64 name: 5-shot - type: accuracy value: 58.85 name: 10-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_gsm8k type: OpenLLM-Ro/ro_gsm8k metrics: - type: accuracy value: 34.77 name: Average accuracy - type: accuracy value: 32.52 name: 1-shot - type: accuracy value: 33.97 name: 3-shot - type: accuracy value: 37.83 name: 5-shot - task: type: text-generation dataset: name: OpenLLM-Ro/ro_truthfulqa type: OpenLLM-Ro/ro_truthfulqa metrics: - type: accuracy value: 41.7 name: Average accuracy - task: type: text-generation dataset: name: LaRoSeDa_binary type: LaRoSeDa_binary metrics: - type: macro-f1 value: 97.48 name: Average macro-f1 - type: macro-f1 value: 97.67 name: 0-shot - type: macro-f1 value: 97.07 name: 1-shot - type: macro-f1 value: 97.4 name: 3-shot - type: macro-f1 value: 97.8 name: 5-shot - task: type: text-generation dataset: name: LaRoSeDa_multiclass type: LaRoSeDa_multiclass metrics: - type: macro-f1 value: 54.0 name: Average macro-f1 - type: macro-f1 value: 58.49 name: 0-shot - type: macro-f1 value: 55.93 name: 1-shot - type: macro-f1 value: 47.7 name: 3-shot - type: macro-f1 value: 53.89 name: 5-shot - task: type: text-generation dataset: name: LaRoSeDa_binary_finetuned type: LaRoSeDa_binary_finetuned metrics: - type: macro-f1 value: 0.0 name: Average macro-f1 - task: type: text-generation dataset: name: LaRoSeDa_multiclass_finetuned type: LaRoSeDa_multiclass_finetuned metrics: - type: macro-f1 value: 0.0 name: Average macro-f1 - task: type: text-generation dataset: name: WMT_EN-RO type: WMT_EN-RO metrics: - type: bleu value: 22.09 name: Average bleu - type: bleu value: 8.63 name: 0-shot - type: bleu value: 25.89 name: 1-shot - type: bleu value: 26.79 name: 3-shot - type: bleu value: 27.05 name: 5-shot - task: type: text-generation dataset: name: WMT_RO-EN type: WMT_RO-EN metrics: - type: bleu value: 23.0 name: Average bleu - type: bleu value: 3.56 name: 0-shot - type: bleu value: 20.66 name: 1-shot - type: bleu value: 33.56 name: 3-shot - type: bleu value: 34.22 name: 5-shot - task: type: text-generation dataset: name: WMT_EN-RO_finetuned type: WMT_EN-RO_finetuned metrics: - type: bleu value: 0.0 name: Average bleu - task: type: text-generation dataset: name: WMT_RO-EN_finetuned type: WMT_RO-EN_finetuned metrics: - type: bleu value: 0.0 name: Average bleu - task: type: text-generation dataset: name: XQuAD type: XQuAD metrics: - type: exact_match value: 26.05 name: Average exact_match - type: f1 value: 42.77 name: Average f1 - task: type: text-generation dataset: name: XQuAD_finetuned type: XQuAD_finetuned metrics: - type: exact_match value: 0.0 name: Average exact_match - type: f1 value: 0.0 name: Average f1 - task: type: text-generation dataset: name: STS type: STS metrics: - type: spearman value: 79.64 name: Average spearman - type: pearson value: 79.52 name: Average pearson - task: type: text-generation dataset: name: STS_finetuned type: STS_finetuned metrics: - type: spearman value: 0.0 name: Average spearman - type: pearson value: 0.0 name: Average pearson - task: type: text-generation dataset: name: XQuAD_EM type: XQuAD_EM metrics: - type: exact_match value: 11.26 name: 0-shot - type: exact_match value: 34.29 name: 1-shot - type: exact_match value: 29.24 name: 3-shot - type: exact_match value: 29.41 name: 5-shot - task: type: text-generation dataset: name: XQuAD_F1 type: XQuAD_F1 metrics: - type: f1 value: 22.98 name: 0-shot - type: f1 value: 54.48 name: 1-shot - type: f1 value: 46.18 name: 3-shot - type: f1 value: 47.43 name: 5-shot - task: type: text-generation dataset: name: STS_Spearman type: STS_Spearman metrics: - type: spearman value: 79.99 name: 1-shot - type: spearman value: 78.42 name: 3-shot - type: spearman value: 80.51 name: 5-shot - task: type: text-generation dataset: name: STS_Pearson type: STS_Pearson metrics: - type: pearson value: 80.59 name: 1-shot - type: pearson value: 78.11 name: 3-shot - type: pearson value: 79.87 name: 5-shot --- # code380/RoLlama3-8b-Instruct-DPO-Q8_0-GGUF This model was converted to GGUF format from [`OpenLLM-Ro/RoLlama3-8b-Instruct-DPO`](https://huggingface.co/OpenLLM-Ro/RoLlama3-8b-Instruct-DPO) 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/RoLlama3-8b-Instruct-DPO) 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 code380/RoLlama3-8b-Instruct-DPO-Q8_0-GGUF --hf-file rollama3-8b-instruct-dpo-q8_0.gguf -p "The meaning to life and the universe is" ``` ### Server: ```bash llama-server --hf-repo code380/RoLlama3-8b-Instruct-DPO-Q8_0-GGUF --hf-file rollama3-8b-instruct-dpo-q8_0.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 code380/RoLlama3-8b-Instruct-DPO-Q8_0-GGUF --hf-file rollama3-8b-instruct-dpo-q8_0.gguf -p "The meaning to life and the universe is" ``` or ``` ./llama-server --hf-repo code380/RoLlama3-8b-Instruct-DPO-Q8_0-GGUF --hf-file rollama3-8b-instruct-dpo-q8_0.gguf -c 2048 ```