| --- |
| library_name: llama.cpp |
| license: gemma |
| tags: [] |
| widget: |
| - text: '<start_of_turn>user |
| |
| How does the brain work?<end_of_turn> |
| |
| <start_of_turn>model |
| |
| ' |
| inference: |
| parameters: |
| max_new_tokens: 200 |
| extra_gated_heading: Access Gemma on Hugging Face |
| extra_gated_prompt: To access Gemma on Hugging Face, you’re required to review and |
| agree to Google’s usage license. To do this, please ensure you’re logged-in to Hugging |
| Face and click below. Requests are processed immediately. |
| extra_gated_button_content: Acknowledge license |
| --- |
| |
| # Gemma Model Card |
|
|
| **Model Page**: [Gemma](https://ai.google.dev/gemma/docs) |
|
|
| This model card corresponds to the 2B instruct version of the Gemma model in GGUF Format. The weights here are **float32**. |
|
|
| > [!IMPORTANT] |
| > |
| > In llama.cpp, and other related tools such as Ollama and LM Studio, please make sure that you have these flags set correctly, especially **`repeat-penalty`**. Georgi Gerganov (llama.cpp's author) shared his experience in https://huggingface.co/google/gemma-7b-it/discussions/38#65d7b14adb51f7c160769fa1. |
|
|
| You can also visit the model card of the [2B base model GGUF](https://huggingface.co/google/gemma-2b-GGUF), [7B base model GGUF](https://huggingface.co/google/gemma-7b-GGUF), and [7B instruct model GGUF](https://huggingface.co/google/gemma-7b-it-GGUF). |
|
|
| **Resources and Technical Documentation**: |
|
|
| * [Responsible Generative AI Toolkit](https://ai.google.dev/responsible) |
| * [Gemma on Kaggle](https://www.kaggle.com/models/google/gemma) |
| * [Gemma on Vertex Model Garden](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/335?version=gemma-7b-it-gg-hf) |
|
|
| **Terms of Use**: [Terms](https://www.kaggle.com/models/google/gemma/license/consent/verify/huggingface?returnModelRepoId=google/gemma-2b-it-GGUF) |
|
|
| **Authors**: Google |
|
|
| ## Model Information |
|
|
| Summary description and brief definition of inputs and outputs. |
|
|
| ### Description |
|
|
| Gemma is a family of lightweight, state-of-the-art open models from Google, |
| built from the same research and technology used to create the Gemini models. |
| They are text-to-text, decoder-only large language models, available in English, |
| with open weights, pre-trained variants, and instruction-tuned variants. Gemma |
| models are well-suited for a variety of text generation tasks, including |
| question answering, summarization, and reasoning. Their relatively small size |
| makes it possible to deploy them in environments with limited resources such as |
| a laptop, desktop or your own cloud infrastructure, democratizing access to |
| state of the art AI models and helping foster innovation for everyone. |
|
|
| ### Usage |
|
|
| Below we share some commands on how to get quickly started with running the model. |
|
|
| #### Running the model on a CPU |
|
|
| ```shell |
| llama.cpp/build$ bin/main -m gemma-2b-it.gguf -n 256 -p "Write a Python function to find sum of two numbers." --repeat-penalty 1.1 |
| Log start |
| main: build = 2249 (15499eb9) |
| main: built with cc (Debian 13.2.0-5) 13.2.0 for x86_64-linux-gnu |
| main: seed = 1708973831 |
| llama_model_loader: loaded meta data with 19 key-value pairs and 164 tensors from gemma-2b-it.gguf (version GGUF V3 (latest)) |
| llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. |
| llama_model_loader: - kv 0: general.architecture str = gemma |
| llama_model_loader: - kv 1: general.name str = gemma-2b-it |
| llama_model_loader: - kv 2: gemma.context_length u32 = 8192 |
| llama_model_loader: - kv 3: gemma.block_count u32 = 18 |
| llama_model_loader: - kv 4: gemma.embedding_length u32 = 2048 |
| llama_model_loader: - kv 5: gemma.feed_forward_length u32 = 16384 |
| llama_model_loader: - kv 6: gemma.attention.head_count u32 = 8 |
| llama_model_loader: - kv 7: gemma.attention.head_count_kv u32 = 1 |
| llama_model_loader: - kv 8: gemma.attention.key_length u32 = 256 |
| llama_model_loader: - kv 9: gemma.attention.value_length u32 = 256 |
| llama_model_loader: - kv 10: gemma.attention.layer_norm_rms_epsilon f32 = 0.000001 |
| llama_model_loader: - kv 11: tokenizer.ggml.model str = llama |
| llama_model_loader: - kv 12: tokenizer.ggml.bos_token_id u32 = 2 |
| llama_model_loader: - kv 13: tokenizer.ggml.eos_token_id u32 = 1 |
| llama_model_loader: - kv 14: tokenizer.ggml.padding_token_id u32 = 0 |
| llama_model_loader: - kv 15: tokenizer.ggml.unknown_token_id u32 = 3 |
| llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,256128] = ["<pad>", "<eos>", "<bos>", "<unk>", ... |
| llama_model_loader: - kv 17: tokenizer.ggml.scores arr[f32,256128] = [0.000000, 0.000000, 0.000000, 0.0000... |
| llama_model_loader: - kv 18: tokenizer.ggml.token_type arr[i32,256128] = [3, 3, 3, 2, 1, 1, 1, 1, 1, 1, 1, 1, ... |
| llama_model_loader: - type f32: 164 tensors |
| llm_load_vocab: mismatch in special tokens definition ( 544/256128 vs 388/256128 ). |
| llm_load_print_meta: format = GGUF V3 (latest) |
| llm_load_print_meta: arch = gemma |
| llm_load_print_meta: vocab type = SPM |
| llm_load_print_meta: n_vocab = 256128 |
| llm_load_print_meta: n_merges = 0 |
| llm_load_print_meta: n_ctx_train = 8192 |
| llm_load_print_meta: n_embd = 2048 |
| llm_load_print_meta: n_head = 8 |
| llm_load_print_meta: n_head_kv = 1 |
| llm_load_print_meta: n_layer = 18 |
| llm_load_print_meta: n_rot = 256 |
| llm_load_print_meta: n_embd_head_k = 256 |
| llm_load_print_meta: n_embd_head_v = 256 |
| llm_load_print_meta: n_gqa = 8 |
| llm_load_print_meta: n_embd_k_gqa = 256 |
| llm_load_print_meta: n_embd_v_gqa = 256 |
| llm_load_print_meta: f_norm_eps = 0.0e+00 |
| llm_load_print_meta: f_norm_rms_eps = 1.0e-06 |
| llm_load_print_meta: f_clamp_kqv = 0.0e+00 |
| llm_load_print_meta: f_max_alibi_bias = 0.0e+00 |
| llm_load_print_meta: n_ff = 16384 |
| llm_load_print_meta: n_expert = 0 |
| llm_load_print_meta: n_expert_used = 0 |
| llm_load_print_meta: rope scaling = linear |
| llm_load_print_meta: freq_base_train = 10000.0 |
| llm_load_print_meta: freq_scale_train = 1 |
| llm_load_print_meta: n_yarn_orig_ctx = 8192 |
| llm_load_print_meta: rope_finetuned = unknown |
| llm_load_print_meta: model type = 2B |
| llm_load_print_meta: model ftype = all F32 (guessed) |
| llm_load_print_meta: model params = 2.51 B |
| llm_load_print_meta: model size = 9.34 GiB (32.00 BPW) |
| llm_load_print_meta: general.name = gemma-2b-it |
| llm_load_print_meta: BOS token = 2 '<bos>' |
| llm_load_print_meta: EOS token = 1 '<eos>' |
| llm_load_print_meta: UNK token = 3 '<unk>' |
| llm_load_print_meta: PAD token = 0 '<pad>' |
| llm_load_print_meta: LF token = 227 '<0x0A>' |
| llm_load_tensors: ggml ctx size = 0.06 MiB |
| llm_load_tensors: CPU buffer size = 9561.29 MiB |
| ............................................................. |
| llama_new_context_with_model: n_ctx = 512 |
| llama_new_context_with_model: freq_base = 10000.0 |
| llama_new_context_with_model: freq_scale = 1 |
| llama_kv_cache_init: CPU KV buffer size = 9.00 MiB |
| llama_new_context_with_model: KV self size = 9.00 MiB, K (f16): 4.50 MiB, V (f16): 4.50 MiB |
| llama_new_context_with_model: CPU input buffer size = 6.01 MiB |
| llama_new_context_with_model: CPU compute buffer size = 504.25 MiB |
| llama_new_context_with_model: graph splits (measure): 1 |
| |
| system_info: n_threads = 24 / 48 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | |
| sampling: |
| repeat_last_n = 64, repeat_penalty = 1.100, frequency_penalty = 0.000, presence_penalty = 0.000 |
| top_k = 40, tfs_z = 1.000, top_p = 0.950, min_p = 0.050, typical_p = 1.000, temp = 0.800 |
| mirostat = 0, mirostat_lr = 0.100, mirostat_ent = 5.000 |
| sampling order: |
| CFG -> Penalties -> top_k -> tfs_z -> typical_p -> top_p -> min_p -> temperature |
| generate: n_ctx = 512, n_batch = 512, n_predict = 256, n_keep = 1 |
| |
| |
| Write a Python function to find sum of two numbers. |
| |
| ```python |
| def sum(a, b): |
| """ |
| Finds the sum of two numbers. |
|
|
| Args: |
| a (int): The first number. |
| b (int): The second number. |
| |
| Returns: |
| int: The sum of a and b. |
| """ |
| |
| return a + b |
|
|
|
|
| # Get the two numbers from the user. |
| num1 = int(input("Enter the first number: ")) |
| num2 = int(input("Enter the second number: ")) |
|
|
| # Find the sum of num1 and num2. |
| sum_of_numbers = sum(num1, num2) |
|
|
| # Print the sum of num1 and num2. |
| print(f"The sum of {num1} and {num2} is {sum_of_numbers}") |
| ``` [end of text] |
| |
| llama_print_timings: load time = 4991.67 ms |
| llama_print_timings: sample time = 201.42 ms / 177 runs ( 1.14 ms per token, 878.76 tokens per second) |
| llama_print_timings: prompt eval time = 222.46 ms / 12 tokens ( 18.54 ms per token, 53.94 tokens per second) |
| llama_print_timings: eval time = 29505.66 ms / 176 runs ( 167.65 ms per token, 5.96 tokens per second) |
| llama_print_timings: total time = 30275.88 ms / 188 tokens |
| Log end |
| ``` |
|
|
|
|
| #### Running the model on a single / multi GPU |
|
|
|
|
| ```shell |
| llama.cpp/build$ bin/main -m gemma-2b-it.gguf -n 256 -p "Write a Python function to find sum of two numbers." --repeat-penalty 1.1 -ngl 99 |
| Log start |
| main: build = 2234 (973053d8) |
| main: built with cc (Debian 13.2.0-5) 13.2.0 for x86_64-linux-gnu |
| main: seed = 1708975775 |
| ggml_init_cublas: GGML_CUDA_FORCE_MMQ: no |
| ggml_init_cublas: CUDA_USE_TENSOR_CORES: yes |
| ggml_init_cublas: found 1 CUDA devices: |
| Device 0: NVIDIA GeForce RTX 3060, compute capability 8.6, VMM: yes |
| llama_model_loader: loaded meta data with 19 key-value pairs and 164 tensors from gemma-2b-it.gguf (version GGUF V3 (latest)) |
| llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. |
| llama_model_loader: - kv 0: general.architecture str = gemma |
| llama_model_loader: - kv 1: general.name str = gemma-2b-it |
| llama_model_loader: - kv 2: gemma.context_length u32 = 8192 |
| llama_model_loader: - kv 3: gemma.block_count u32 = 18 |
| llama_model_loader: - kv 4: gemma.embedding_length u32 = 2048 |
| llama_model_loader: - kv 5: gemma.feed_forward_length u32 = 16384 |
| llama_model_loader: - kv 6: gemma.attention.head_count u32 = 8 |
| llama_model_loader: - kv 7: gemma.attention.head_count_kv u32 = 1 |
| llama_model_loader: - kv 8: gemma.attention.key_length u32 = 256 |
| llama_model_loader: - kv 9: gemma.attention.value_length u32 = 256 |
| llama_model_loader: - kv 10: gemma.attention.layer_norm_rms_epsilon f32 = 0.000001 |
| llama_model_loader: - kv 11: tokenizer.ggml.model str = llama |
| llama_model_loader: - kv 12: tokenizer.ggml.bos_token_id u32 = 2 |
| llama_model_loader: - kv 13: tokenizer.ggml.eos_token_id u32 = 1 |
| llama_model_loader: - kv 14: tokenizer.ggml.padding_token_id u32 = 0 |
| llama_model_loader: - kv 15: tokenizer.ggml.unknown_token_id u32 = 3 |
| llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,256128] = ["<pad>", "<eos>", "<bos>", "<unk>", ... |
| llama_model_loader: - kv 17: tokenizer.ggml.scores arr[f32,256128] = [0.000000, 0.000000, 0.000000, 0.0000... |
| llama_model_loader: - kv 18: tokenizer.ggml.token_type arr[i32,256128] = [3, 3, 3, 2, 1, 1, 1, 1, 1, 1, 1, 1, ... |
| llama_model_loader: - type f32: 164 tensors |
| llm_load_vocab: mismatch in special tokens definition ( 544/256128 vs 388/256128 ). |
| llm_load_print_meta: format = GGUF V3 (latest) |
| llm_load_print_meta: arch = gemma |
| llm_load_print_meta: vocab type = SPM |
| llm_load_print_meta: n_vocab = 256128 |
| llm_load_print_meta: n_merges = 0 |
| llm_load_print_meta: n_ctx_train = 8192 |
| llm_load_print_meta: n_embd = 2048 |
| llm_load_print_meta: n_head = 8 |
| llm_load_print_meta: n_head_kv = 1 |
| llm_load_print_meta: n_layer = 18 |
| llm_load_print_meta: n_rot = 256 |
| llm_load_print_meta: n_embd_head_k = 256 |
| llm_load_print_meta: n_embd_head_v = 256 |
| llm_load_print_meta: n_gqa = 8 |
| llm_load_print_meta: n_embd_k_gqa = 256 |
| llm_load_print_meta: n_embd_v_gqa = 256 |
| llm_load_print_meta: f_norm_eps = 0.0e+00 |
| llm_load_print_meta: f_norm_rms_eps = 1.0e-06 |
| llm_load_print_meta: f_clamp_kqv = 0.0e+00 |
| llm_load_print_meta: f_max_alibi_bias = 0.0e+00 |
| llm_load_print_meta: n_ff = 16384 |
| llm_load_print_meta: n_expert = 0 |
| llm_load_print_meta: n_expert_used = 0 |
| llm_load_print_meta: rope scaling = linear |
| llm_load_print_meta: freq_base_train = 10000.0 |
| llm_load_print_meta: freq_scale_train = 1 |
| llm_load_print_meta: n_yarn_orig_ctx = 8192 |
| llm_load_print_meta: rope_finetuned = unknown |
| llm_load_print_meta: model type = 2B |
| llm_load_print_meta: model ftype = all F32 (guessed) |
| llm_load_print_meta: model params = 2.51 B |
| llm_load_print_meta: model size = 9.34 GiB (32.00 BPW) |
| llm_load_print_meta: general.name = gemma-2b-it |
| llm_load_print_meta: BOS token = 2 '<bos>' |
| llm_load_print_meta: EOS token = 1 '<eos>' |
| llm_load_print_meta: UNK token = 3 '<unk>' |
| llm_load_print_meta: PAD token = 0 '<pad>' |
| llm_load_print_meta: LF token = 227 '<0x0A>' |
| llm_load_tensors: ggml ctx size = 0.13 MiB |
| llm_load_tensors: offloading 18 repeating layers to GPU |
| llm_load_tensors: offloading non-repeating layers to GPU |
| llm_load_tensors: offloaded 19/19 layers to GPU |
| llm_load_tensors: CPU buffer size = 2001.00 MiB |
| llm_load_tensors: CUDA0 buffer size = 9561.29 MiB |
| ............................................................. |
| llama_new_context_with_model: n_ctx = 512 |
| llama_new_context_with_model: freq_base = 10000.0 |
| llama_new_context_with_model: freq_scale = 1 |
| llama_kv_cache_init: CUDA0 KV buffer size = 9.00 MiB |
| llama_new_context_with_model: KV self size = 9.00 MiB, K (f16): 4.50 MiB, V (f16): 4.50 MiB |
| llama_new_context_with_model: CUDA_Host input buffer size = 6.01 MiB |
| ggml_gallocr_reserve_n: reallocating CUDA0 buffer from size 0.00 MiB to 508.25 MiB |
| ggml_gallocr_reserve_n: reallocating CUDA_Host buffer from size 0.00 MiB to 4.00 MiB |
| llama_new_context_with_model: CUDA0 compute buffer size = 508.25 MiB |
| llama_new_context_with_model: CUDA_Host compute buffer size = 4.00 MiB |
| llama_new_context_with_model: graph splits (measure): 3 |
| ggml_gallocr_needs_realloc: graph has different number of nodes |
| ggml_gallocr_alloc_graph: cannot reallocate multi buffer graph automatically, call reserve |
| ggml_backend_sched: failed to allocate graph, reserving |
| |
| system_info: n_threads = 6 / 12 | AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | FMA = 1 | NEON = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | |
| sampling: |
| repeat_last_n = 64, repeat_penalty = 1.100, frequency_penalty = 0.000, presence_penalty = 0.000 |
| top_k = 40, tfs_z = 1.000, top_p = 0.950, min_p = 0.050, typical_p = 1.000, temp = 0.800 |
| mirostat = 0, mirostat_lr = 0.100, mirostat_ent = 5.000 |
| sampling order: |
| CFG -> Penalties -> top_k -> tfs_z -> typical_p -> top_p -> min_p -> temperature |
| generate: n_ctx = 512, n_batch = 512, n_predict = 256, n_keep = 1 |
| |
| |
| Write a Python function to find sum of two numbers.ggml_gallocr_needs_realloc: node inp_embd is not valid |
| ggml_gallocr_alloc_graph: cannot reallocate multi buffer graph automatically, call reserve |
| ggml_backend_sched: failed to allocate graph, reserving |
| |
| |
| ```python |
| def sum(a, b): |
| return a + b |
| |
|
|
| # Get the two numbers from the user |
| num1 = int(input("Enter first number: ")) |
| num2 = int(input("Enter second number: ")) |
|
|
| # Calculate and print the sum |
| sum_of_numbers = sum(num1, num2) |
| print(f"Sum of {num1} and {num2} is {sum_of_numbers}") |
| ``` [end of text] |
| |
| llama_print_timings: load time = 12672.17 ms |
| llama_print_timings: sample time = 2352.84 ms / 101 runs ( 23.30 ms per token, 42.93 tokens per second) |
| llama_print_timings: prompt eval time = 35.28 ms / 12 tokens ( 2.94 ms per token, 340.16 tokens per second) |
| llama_print_timings: eval time = 3189.28 ms / 100 runs ( 31.89 ms per token, 31.36 tokens per second) |
| llama_print_timings: total time = 5898.47 ms / 112 tokens |
| Log end |
| ``` |
|
|
| ### Inputs and outputs |
|
|
| * **Input:** Text string, such as a question, a prompt, or a document to be |
| summarized. |
| * **Output:** Generated English-language text in response to the input, such |
| as an answer to a question, or a summary of a document. |
| |
| ## Model Data |
|
|
| Data used for model training and how the data was processed. |
|
|
| ### Training Dataset |
|
|
| These models were trained on a dataset of text data that includes a wide variety |
| of sources, totaling 6 trillion tokens. Here are the key components: |
|
|
| * Web Documents: A diverse collection of web text ensures the model is exposed |
| to a broad range of linguistic styles, topics, and vocabulary. Primarily |
| English-language content. |
| * Code: Exposing the model to code helps it to learn the syntax and patterns of |
| programming languages, which improves its ability to generate code or |
| understand code-related questions. |
| * Mathematics: Training on mathematical text helps the model learn logical |
| reasoning, symbolic representation, and to address mathematical queries. |
|
|
| The combination of these diverse data sources is crucial for training a powerful |
| language model that can handle a wide variety of different tasks and text |
| formats. |
|
|
| ### Data Preprocessing |
|
|
| Here are the key data cleaning and filtering methods applied to the training |
| data: |
|
|
| * CSAM Filtering: Rigorous CSAM (Child Sexual Abuse Material) filtering was |
| applied at multiple stages in the data preparation process to ensure the |
| exclusion of harmful and illegal content |
| * Sensitive Data Filtering: As part of making Gemma pre-trained models safe and |
| reliable, automated techniques were used to filter out certain personal |
| information and other sensitive data from training sets. |
| * Additional methods: Filtering based on content quality and safely in line with |
| [our policies](https://storage.googleapis.com/gweb-uniblog-publish-prod/documents/2023_Google_AI_Principles_Progress_Update.pdf#page=11). |
|
|
| ## Implementation Information |
|
|
| Details about the model internals. |
|
|
| ### Hardware |
|
|
| Gemma was trained using the latest generation of |
| [Tensor Processing Unit (TPU)](https://cloud.google.com/tpu/docs/intro-to-tpu) hardware (TPUv5e). |
|
|
| Training large language models requires significant computational power. TPUs, |
| designed specifically for matrix operations common in machine learning, offer |
| several advantages in this domain: |
|
|
| * Performance: TPUs are specifically designed to handle the massive computations |
| involved in training LLMs. They can speed up training considerably compared to |
| CPUs. |
| * Memory: TPUs often come with large amounts of high-bandwidth memory, allowing |
| for the handling of large models and batch sizes during training. This can |
| lead to better model quality. |
| * Scalability: TPU Pods (large clusters of TPUs) provide a scalable solution for |
| handling the growing complexity of large foundation models. You can distribute |
| training across multiple TPU devices for faster and more efficient processing. |
| * Cost-effectiveness: In many scenarios, TPUs can provide a more cost-effective |
| solution for training large models compared to CPU-based infrastructure, |
| especially when considering the time and resources saved due to faster |
| training. |
| * These advantages are aligned with |
| [Google's commitments to operate sustainably](https://sustainability.google/operating-sustainably/). |
|
|
| ### Software |
|
|
| Training was done using [JAX](https://github.com/google/jax) and [ML Pathways](https://blog.google/technology/ai/introducing-pathways-next-generation-ai-architecture). |
|
|
| JAX allows researchers to take advantage of the latest generation of hardware, |
| including TPUs, for faster and more efficient training of large models. |
|
|
| ML Pathways is Google's latest effort to build artificially intelligent systems |
| capable of generalizing across multiple tasks. This is specially suitable for |
| [foundation models](https://ai.google/discover/foundation-models/), including large language models like |
| these ones. |
|
|
| Together, JAX and ML Pathways are used as described in the |
| [paper about the Gemini family of models](https://arxiv.org/abs/2312.11805); "the 'single |
| controller' programming model of Jax and Pathways allows a single Python |
| process to orchestrate the entire training run, dramatically simplifying the |
| development workflow." |
|
|
| ## Evaluation |
|
|
| Model evaluation metrics and results. |
|
|
| ### Benchmark Results |
|
|
| These models were evaluated against a large collection of different datasets and |
| metrics to cover different aspects of text generation: |
|
|
| | Benchmark | Metric | 2B Params | 7B Params | |
| | ------------------------------ | ------------- | ----------- | --------- | |
| | [MMLU](https://arxiv.org/abs/2009.03300) | 5-shot, top-1 | 42.3 | 64.3 | |
| | [HellaSwag](https://arxiv.org/abs/1905.07830) | 0-shot |71.4 | 81.2 | |
| | [PIQA](https://arxiv.org/abs/1911.11641) | 0-shot | 77.3 | 81.2 | |
| | [SocialIQA](https://arxiv.org/abs/1904.09728) | 0-shot | 59.7 | 51.8 | |
| | [BooIQ](https://arxiv.org/abs/1905.10044) | 0-shot | 69.4 | 83.2 | |
| | [WinoGrande](https://arxiv.org/abs/1907.10641) | partial score | 65.4 | 72.3 | |
| | [CommonsenseQA](https://arxiv.org/abs/1811.00937) | 7-shot | 65.3 | 71.3 | |
| | [OpenBookQA](https://arxiv.org/abs/1809.02789) | | 47.8 | 52.8 | |
| | [ARC-e](https://arxiv.org/abs/1911.01547) | | 73.2 | 81.5 | |
| | [ARC-c](https://arxiv.org/abs/1911.01547) | | 42.1 | 53.2 | |
| | [TriviaQA](https://arxiv.org/abs/1705.03551) | 5-shot | 53.2 | 63.4 | |
| | [Natural Questions](https://github.com/google-research-datasets/natural-questions) | 5-shot | - | 23 | |
| | [HumanEval](https://arxiv.org/abs/2107.03374) | pass@1 | 22.0 | 32.3 | |
| | [MBPP](https://arxiv.org/abs/2108.07732) | 3-shot | 29.2 | 44.4 | |
| | [GSM8K](https://arxiv.org/abs/2110.14168) | maj@1 | 17.7 | 46.4 | |
| | [MATH](https://arxiv.org/abs/2108.07732) | 4-shot | 11.8 | 24.3 | |
| | [AGIEval](https://arxiv.org/abs/2304.06364) | | 24.2 | 41.7 | |
| | [BIG-Bench](https://arxiv.org/abs/2206.04615) | | 35.2 | 55.1 | |
| | ------------------------------ | ------------- | ----------- | --------- | |
| | **Average** | | **54.0** | **56.4** | |
|
|
| ## Ethics and Safety |
|
|
| Ethics and safety evaluation approach and results. |
|
|
| ### Evaluation Approach |
|
|
| Our evaluation methods include structured evaluations and internal red-teaming |
| testing of relevant content policies. Red-teaming was conducted by a number of |
| different teams, each with different goals and human evaluation metrics. These |
| models were evaluated against a number of different categories relevant to |
| ethics and safety, including: |
|
|
| * Text-to-Text Content Safety: Human evaluation on prompts covering safety |
| policies including child sexual abuse and exploitation, harassment, violence |
| and gore, and hate speech. |
| * Text-to-Text Representational Harms: Benchmark against relevant academic |
| datasets such as [WinoBias](https://arxiv.org/abs/1804.06876) and [BBQ Dataset](https://arxiv.org/abs/2110.08193v2). |
| * Memorization: Automated evaluation of memorization of training data, including |
| the risk of personally identifiable information exposure. |
| * Large-scale harm: Tests for "dangerous capabilities," such as chemical, |
| biological, radiological, and nuclear (CBRN) risks. |
|
|
| ### Evaluation Results |
|
|
| The results of ethics and safety evaluations are within acceptable thresholds |
| for meeting [internal policies](https://storage.googleapis.com/gweb-uniblog-publish-prod/documents/2023_Google_AI_Principles_Progress_Update.pdf#page=11) for categories such as child |
| safety, content safety, representational harms, memorization, large-scale harms. |
| On top of robust internal evaluations, the results of well known safety |
| benchmarks like BBQ, BOLD, Winogender, Winobias, RealToxicity, and TruthfulQA |
| are shown here. |
|
|
| | Benchmark | Metric | 2B Params | 7B Params | |
| | ------------------------------ | ------------- | ----------- | --------- | |
| | [RealToxicity](https://arxiv.org/abs/2009.11462) | average | 6.86 | 7.90 | |
| | [BOLD](https://arxiv.org/abs/2101.11718) | | 45.57 | 49.08 | |
| | [CrowS-Pairs](https://aclanthology.org/2020.emnlp-main.154/) | top-1 | 45.82 | 51.33 | |
| | [BBQ Ambig](https://arxiv.org/abs/2110.08193v2) | 1-shot, top-1 | 62.58 | 92.54 | |
| | [BBQ Disambig](https://arxiv.org/abs/2110.08193v2) | top-1 | 54.62 | 71.99 | |
| | [Winogender](https://arxiv.org/abs/1804.09301) | top-1 | 51.25 | 54.17 | |
| | [TruthfulQA](https://arxiv.org/abs/2109.07958) | | 44.84 | 31.81 | |
| | [Winobias 1_2](https://arxiv.org/abs/1804.06876) | | 56.12 | 59.09 | |
| | [Winobias 2_2](https://arxiv.org/abs/1804.06876) | | 91.10 | 92.23 | |
| | [Toxigen](https://arxiv.org/abs/2203.09509) | | 29.77 | 39.59 | |
| | ------------------------------ | ------------- | ----------- | --------- | |
|
|
|
|
| ## Usage and Limitations |
|
|
| These models have certain limitations that users should be aware of. |
|
|
| ### Intended Usage |
|
|
| Open Large Language Models (LLMs) have a wide range of applications across |
| various industries and domains. The following list of potential uses is not |
| comprehensive. The purpose of this list is to provide contextual information |
| about the possible use-cases that the model creators considered as part of model |
| training and development. |
|
|
| * Content Creation and Communication |
| * Text Generation: These models can be used to generate creative text formats |
| such as poems, scripts, code, marketing copy, and email drafts. |
| * Chatbots and Conversational AI: Power conversational interfaces for customer |
| service, virtual assistants, or interactive applications. |
| * Text Summarization: Generate concise summaries of a text corpus, research |
| papers, or reports. |
| * Research and Education |
| * Natural Language Processing (NLP) Research: These models can serve as a |
| foundation for researchers to experiment with NLP techniques, develop |
| algorithms, and contribute to the advancement of the field. |
| * Language Learning Tools: Support interactive language learning experiences, |
| aiding in grammar correction or providing writing practice. |
| * Knowledge Exploration: Assist researchers in exploring large bodies of text |
| by generating summaries or answering questions about specific topics. |
| |
| ### Limitations |
|
|
| * Training Data |
| * The quality and diversity of the training data significantly influence the |
| model's capabilities. Biases or gaps in the training data can lead to |
| limitations in the model's responses. |
| * The scope of the training dataset determines the subject areas the model can |
| handle effectively. |
| * Context and Task Complexity |
| * LLMs are better at tasks that can be framed with clear prompts and |
| instructions. Open-ended or highly complex tasks might be challenging. |
| * A model's performance can be influenced by the amount of context provided |
| (longer context generally leads to better outputs, up to a certain point). |
| * Language Ambiguity and Nuance |
| * Natural language is inherently complex. LLMs might struggle to grasp subtle |
| nuances, sarcasm, or figurative language. |
| * Factual Accuracy |
| * LLMs generate responses based on information they learned from their |
| training datasets, but they are not knowledge bases. They may generate |
| incorrect or outdated factual statements. |
| * Common Sense |
| * LLMs rely on statistical patterns in language. They might lack the ability |
| to apply common sense reasoning in certain situations. |
| |
| ### Ethical Considerations and Risks |
|
|
| The development of large language models (LLMs) raises several ethical concerns. |
| In creating an open model, we have carefully considered the following: |
|
|
| * Bias and Fairness |
| * LLMs trained on large-scale, real-world text data can reflect socio-cultural |
| biases embedded in the training material. These models underwent careful |
| scrutiny, input data pre-processing described and posterior evaluations |
| reported in this card. |
| * Misinformation and Misuse |
| * LLMs can be misused to generate text that is false, misleading, or harmful. |
| * Guidelines are provided for responsible use with the model, see the |
| [Responsible Generative AI Toolkit](http://ai.google.dev/gemma/responsible). |
| * Transparency and Accountability: |
| * This model card summarizes details on the models' architecture, |
| capabilities, limitations, and evaluation processes. |
| * A responsibly developed open model offers the opportunity to share |
| innovation by making LLM technology accessible to developers and researchers |
| across the AI ecosystem. |
| |
| Risks identified and mitigations: |
|
|
| * Perpetuation of biases: It's encouraged to perform continuous monitoring |
| (using evaluation metrics, human review) and the exploration of de-biasing |
| techniques during model training, fine-tuning, and other use cases. |
| * Generation of harmful content: Mechanisms and guidelines for content safety |
| are essential. Developers are encouraged to exercise caution and implement |
| appropriate content safety safeguards based on their specific product policies |
| and application use cases. |
| * Misuse for malicious purposes: Technical limitations and developer and |
| end-user education can help mitigate against malicious applications of LLMs. |
| Educational resources and reporting mechanisms for users to flag misuse are |
| provided. Prohibited uses of Gemma models are outlined in the |
| [Gemma Prohibited Use Policy](https://ai.google.dev/gemma/prohibited_use_policy). |
| * Privacy violations: Models were trained on data filtered for removal of PII |
| (Personally Identifiable Information). Developers are encouraged to adhere to |
| privacy regulations with privacy-preserving techniques. |
|
|
| ### Benefits |
|
|
| At the time of release, this family of models provides high-performance open |
| large language model implementations designed from the ground up for Responsible |
| AI development compared to similarly sized models. |
|
|
| Using the benchmark evaluation metrics described in this document, these models |
| have shown to provide superior performance to other, comparably-sized open model |
| alternatives. |
|
|
