Quantization made by Richard Erkhov. [Github](https://github.com/RichardErkhov) [Discord](https://discord.gg/pvy7H8DZMG) [Request more models](https://github.com/RichardErkhov/quant_request) Llama-2-7b-dolphin-open_platypus-pruned_70 - GGUF - Model creator: https://huggingface.co/neuralmagic/ - Original model: https://huggingface.co/neuralmagic/Llama-2-7b-dolphin-open_platypus-pruned_70/ | Name | Quant method | Size | | ---- | ---- | ---- | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q2_K.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q2_K.gguf) | Q2_K | 2.36GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.IQ3_XS.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.IQ3_XS.gguf) | IQ3_XS | 2.6GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.IQ3_S.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.IQ3_S.gguf) | IQ3_S | 2.75GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q3_K_S.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q3_K_S.gguf) | Q3_K_S | 2.75GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.IQ3_M.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.IQ3_M.gguf) | IQ3_M | 2.9GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q3_K.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q3_K.gguf) | Q3_K | 3.07GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q3_K_M.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q3_K_M.gguf) | Q3_K_M | 3.07GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q3_K_L.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q3_K_L.gguf) | Q3_K_L | 3.35GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.IQ4_XS.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.IQ4_XS.gguf) | IQ4_XS | 3.4GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_0.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_0.gguf) | Q4_0 | 3.56GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.IQ4_NL.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.IQ4_NL.gguf) | IQ4_NL | 3.58GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_K_S.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_K_S.gguf) | Q4_K_S | 3.59GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_K.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_K.gguf) | Q4_K | 3.8GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_K_M.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_K_M.gguf) | Q4_K_M | 3.8GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_1.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q4_1.gguf) | Q4_1 | 3.95GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_0.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_0.gguf) | Q5_0 | 4.33GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_K_S.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_K_S.gguf) | Q5_K_S | 4.33GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_K.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_K.gguf) | Q5_K | 4.45GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_K_M.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_K_M.gguf) | Q5_K_M | 4.45GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_1.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q5_1.gguf) | Q5_1 | 4.72GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q6_K.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q6_K.gguf) | Q6_K | 5.15GB | | [Llama-2-7b-dolphin-open_platypus-pruned_70.Q8_0.gguf](https://huggingface.co/RichardErkhov/neuralmagic_-_Llama-2-7b-dolphin-open_platypus-pruned_70-gguf/blob/main/Llama-2-7b-dolphin-open_platypus-pruned_70.Q8_0.gguf) | Q8_0 | 6.67GB | Original model description: --- base_model: neuralmagic/Llama-2-7b-pruned70-retrained inference: true model_type: llama pipeline_tag: text-generation datasets: - garage-bAInd/Open-Platypus - Open-Orca/OpenOrca - cognitivecomputations/dolphin tags: - sparse - instruct --- # Llama-2-7b-pruned70-retrained-instruct This repo contains a [70% sparse Llama 2 7B](https://huggingface.co/neuralmagic/Llama-2-7b-pruned70-retrained) finetuned for instruction-following tasks using a blend of the Platypus + Open Orca + Dolphin datasets. Official model weights from [Enabling High-Sparsity Foundational Llama Models with Efficient Pretraining and Deployment](https://arxiv.org/abs/2405.03594). **Authors**: Neural Magic, Cerebras ## Usage Below we share some code snippets on how to get quickly started with running the model. ### Sparse Transfer By leveraging a pre-sparsified model's structure, you can efficiently fine-tune on new data, leading to reduced hyperparameter tuning, training times, and computational costs. Learn about this process [here](https://neuralmagic.github.io/docs-v2/get-started/transfer). ### Running the model This model may be run with the transformers library. For accelerated inference with sparsity, deploy with [nm-vllm](https://github.com/neuralmagic/nm-vllm) or [deepsparse](https://github.com/neuralmagic/deepsparse). ```python # pip install transformers accelerate from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("Llama-2-7b-pruned70-retrained-instruct") model = AutoModelForCausalLM.from_pretrained("Llama-2-7b-pruned70-retrained-instruct", device_map="auto") input_text = "Write a recipe for banana bread:\n" input_ids = tokenizer(input_text, return_tensors="pt").to("cuda") outputs = model.generate(**input_ids) print(tokenizer.decode(outputs[0])) ``` ## Evaluation Benchmark Results Model evaluation metrics and results. | Benchmark | Metric | Llama-2-7b-instruct | Llama-2-7b-pruned70-retrained-instruct | |------------------------------------------------|---------------|-------------|-------------------------------| | [MMLU](https://arxiv.org/abs/2009.03300) | 5-shot | 48.60% | 42.33% | | [HellaSwag](https://arxiv.org/abs/1905.07830) | 10-shot | 79.45% | 77.21% | | [WinoGrande](https://arxiv.org/abs/1907.10641) | 5-shot | 75.69% | 71.90% | | [ARC-c](https://arxiv.org/abs/1911.01547) | 25-shot | 53.92% | 47.35% | | [TruthfulQA](https://arxiv.org/abs/2109.07958) | 0-shot | 43.63% | 42.25% | | [GSM8K](https://arxiv.org/abs/2110.14168) | 5-shot | 15.92% | 14.25% | ## Model Training Details This model was obtained by sparse-tranfer of the sparse foundational model [Llama-2-7b-pruned50-retrained](https://huggingface.co/neuralmagic/Llama-2-7b-pruned70-retrained) on a blend of [Open Platypus](https://huggingface.co/datasets/garage-bAInd/Open-Platypus), 10% [Open Orca](https://huggingface.co/datasets/Open-Orca/OpenOrca) and 10% [Dolphin](https://huggingface.co/datasets/cognitivecomputations/dolphin) datasets. Training was perfomerd for 6 epochs. ## Help For further support, and discussions on these models and AI in general, join [Neural Magic's Slack Community](https://join.slack.com/t/discuss-neuralmagic/shared_invite/zt-q1a1cnvo-YBoICSIw3L1dmQpjBeDurQ)